alternative to configurations

[Gach74]

Good morning.

For company reasons, previous tests and advice from the same software provider (at the moment we are not in service) we need to create variants of some details without resorting to configurations.

are there alternative solutions, even the most arzigogular?
all then should be managed by a db

for clarity idic classic example of particular to manage:
- we imagine a simple tubular with welded two plates without processing at the extremes;
- we manage the thing like 3 separate codes that are eventually conveyed into a code together which is then what ends in the bom
- my problem is the management completely for the variants of the tubular and the axieme (tub+plate) both in the initial version and in the time when to make any modification to the tubular, rather than to the plates.

I hope I've been quite clear in the request and for now I thank those who want to answer

♪

 

[MassiVonWeizen]

Save the bodies in a new file.
this keeps the connection with the original part but obviously losing the function shaft.

I'm still waiting for your feedback on your discussion

Equazione con parte sospesa

Ciao a tutti e grazie anticipato per chi volesse rispondere. Ho questo problema. Creo un assieme di rullo+moizzi, poi lo tornisco ad un diametro variabile volta per volta (nell'ordine di +/- 0.5 mm) e sopra di esso vado ad applicare una parte nell'assieme, quindi non salvata, consistente in un...

www.cad3d.it

 

[Gach74]

Meanwhile thank you for the answer.
I was convinced I had already thanked you, but obviously I remembered badly... so thank you, your solution works.

my 3 details are already 3 separate codes, as the plates are used for n assemblies.
put it more explicitly:
- plate dx -> pipe
- plate sx -> pluto
- tubular -> duck/1000 where 1000 is length
- together (pluto+paperino/1000) = duck/1000/com that is the code that goes in separate

in version 2d the design had code:
duck/a/com where a= all possible variants
and clearly all its tubulars were duck/a.

Now, with the 2d system, when I had to make a change to the final piece .../com I opened the design and signaled the modification in the cartilage and everything ended there.
with the modeler how do I know, when I open any variant, which I should possibly update it?
I clearly thought of creating a master model called duck/a/com that should command all variants...but I don't know how to do

Thanks again
♪

 

[MassiVonWeizen]

I didn't understand anything.
starting from the assumption, to facilitate the speech, that you change only the length of the tube

Now, with the 2d system, when I had to make a change to the final piece .../com I opened the design and signaled the modification in the cartilage and everything ended there.

What does that mean? how did you change the length? write in the cartiglio: rev8 length 500; rev 9 length 600.. .
who opens the drawing always and only the length of the last revision?

with the modeler how do I know, when I open any variant, which I should possibly update it?

What do you mean, physically, by variant? and why do you have to update it?

Do you have to keep all variants or do you have to use only the latest version?

what you want to get to me is very nebulous

I was convinced I had already thanked you, but obviously I remembered badly... so thank you, your solution works.

more than thank you need feedback if a solution has been useful or not because who will happen on the discussion with a similar problem already knows whether the solution has worked or not. report it to the discussion please

 

[Gach74]

Okay, I'll explain.
- with variant I mean a change of length 1000, 1520, 1600 and are all active because they are mounted on different machines or inserted in different groups
- the modification of the variant can be a trivial hole in the tubular or a modification of one of the 2 plates (evenly adding a hole in the same or in one of them)

I try to make an evolving code list
50000/1000/com
∟ 50000/1000
) 50001 (dx plate)
) 50002 (sx plate)
50000/1500/
∟ 50000/1500
) 50001 (dx plate)
) 50002 (sx plate)
50000/0800/com
∟ 50000/0800
) 50001 (dx plate)
) 50002 (sx plate)
now I go to make a change to the tubular that will become in the canonical form:
50000-a/a with -a that represents the variant to that from now on will have to be used for all variants.
so if you need the variant .../1000 everything becomes:
50000-a/1000/com
∟ 50000-a/1000
) 50001 (dx plate)
) 50002 (sx plate)
etc.

the same for the modification of a plate always for the variant 1000:

50000-b/1000/com
∟ 50000-a/1000
) 50001-a (dx plate)
) 50002 (sx plate)

or both in different times:

50000-c/1000/com
∟ 50000-a/1000
) 50001-a (dx plate)
) 50002-a (sx plate)

I hope I have spread as much fog as possible

Thank you.

 

[320i S]

I understand, that the revision is on the file name, correct?

 

[Gach74]

precisely.

 

[320i S]

on this then I find it difficult to continue using the multibody and even worse configurations, if you are without a pdm, a backup copy of the project must be made, before the revision to have a historian.

 

[Gach74]

the hazelnut is the pdm!

when I go to review -a of a particular content in the assemblies how does it behave?

50001 (dx plate) --> 50001-a

And here I think there's no problem.

when I open the assembling 50000/1000/com what does it do?
- charge the saved version then 50001
- upload the updated version 50001-a and then create (or ask to create) automatically the version -a of the .../com
- upload the updated version and overwrite the .../com (I hope not)

Thank you.

 

[MassiVonWeizen]

if the new part takes new name with suffix of the variant and you do not want to use configurations I see 2 roads:
1:
as already suggested to make a default part and from this, by saving bodies, derive all variants.
2:
create a set, if you just don't want to use the multibody, with virtual parts that can then be created with internal references and always tied to that specific set

 

[Gach74]

Never use virtual parts, now I study them a bit and try to see if it can be interesting.

I'm very curious about pdm behavior since we'll have to integrate it soon and so all is what we'll have to handle there.

Thanks I try and get back

 

[Gach74]

I have tried the virtual parts and do not do to our case, as they become new components, I lose all the history of the parts.

I had thought of a solution (taking away from the functioning of ptc creo).

- I create the 50000/a file as a combination where I insert a tubular length greater than the longer variant
- insert in the axieme also a hidden part called 50000/ctrl type a cube 1x1x1
I create all the functions I'll need in this together.

when I have to create a variant type 50000/1000
except in copy 50000/a and change the variant quota (in this case the length)

Now, if I have to change the tubular, inserting a trivial hole I go to save in copy the 50000/a as 50000-a/a and I do the same with the 50000/ctrl calling it 50000-a/ctrl, going to write before doing save by name in 50000/ctrl in a field of customized properties modified.

at this point all the variants of the 50000/... that were already created when they are opened, and in them the bom with the column where I went to write, will show me the modified ert and then the user will know that that variant of part/assieme will have to be obtained by making a save with the name of a later version of the 50000/a file.

tested and perfectly working technique though quite laborious, but working that is the only thing that matters.

I didn't want to expose it immediately because I was looking for alternative solutions and maybe more slender than mine.

Now, my bran lies in how the pdm will behave.

thank you for the answers you will give me

 

[320i S]

pdms do not reason all in the same way, they are programs that go "cucited" on business processes and therefore often a pdm turns out more effective than another.

In my view, the pdm must be chosen on aspects such as this (historical and its management).

in pdms usually, replacing codes (because in your case it is the entire code to be replaced, it is not a real revision of the code), it can be made massive or partial, therefore, pdms keep track of everything (including geometric references to other parts/codes even if not related to the project), it is up to you to decide what to show or not.

from me the pdm does not manage the historian to 100% through the revisions, so when we want to track an important change, we change code, the previous one is put obsolete, with tracement of which code replaced it; the revisions, concern only small changes that allow the interchangeability of 100% of the revised code and that the differences, do not affect the concept of the piece itself.

 

[320i S]

tested and perfectly working technique though quite laborious, but working that is the only thing that matters.

attention that the "laborious" techniques always have several faults, before all simplicity and however must guarantee the historian as solid as possible (at least 5/10 years back).

 

[Gach74]

Sorry. 320i s Can I ask which pdm you use?

this laborious solution has been tested and counter-tested and guarantees the historian even beyond the life of the carrier code, the only flaw is the failure to create (remember to create) the new version of .../ctrl.

Unfortunately, as you can understand from this example, the internal management here is complicated and forward there are pieces that also have 1000 variants, which would be 1000 moltriplicate configurations for the number of revisions.

here is the reason for the search for an alternative solution that turns around the problem.

 

[320i S]

I used ptc winchills on pro/e - creo, enovia on catia v6 and currently rule desgin on solidworks

 

[Gach74]

i.speravo i told me epdm the native one of solidworks

 

[320i S]

Sorry. 320i s Can I ask which pdm you use?

this laborious solution has been tested and counter-tested and guarantees the historian even beyond the life of the carrier code, the only flaw is the failure to create (remember to create) the new version of .../ctrl.

Unfortunately, as you can understand from this example, the internal management here is complicated and forward there are pieces that also have 1000 variants, which would be 1000 moltriplicate configurations for the number of revisions.

here is the reason for the search for an alternative solution that turns around the problem.

it would be to be understood if it really makes sense to manage certain details as "integrated to drawings of together", rather than as individual elements, many times chosen of "faciility" of some corporate bodies, they cognize a lot with the management of other agencies (and let's talk to us clearly, almost always does the technical office), the wrong thing is to think about stagnant compartments and not companyly.

 

[320i S]

i.speravo i told me epdm the native one of solidworks

I don't know this, but don't think that because it's done and integrated into the program, it's the only possible alternative. . .

 

[Gach74]

management as explained in fact comes from an imposition of the production office even if it does not find me so disconnected.
However, neglecting the fact of the plates+tubolare also the tubular only must be managed for the various lengths and unfortunately the software has proved lacking in the management variants through configurations.

clearly the problems were found on much more complex details of the tubular.