topic Re: Regeneration errors in Merged Parts/Assemblies? in 3D Part & Assembly Design

Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Fri, 04 May 2018 21:21:01 GMT

For many years my company has been using a merged part in order to simplify and represent an actual manufactured Sub-Assembly (S/A). The problem we are having with this is when one of the base parts change the merged part does not update unless it is checked out and manually regenerated.

Re: Regeneration errors in Merged Parts/Assemblies?

BrianMartin — Thu, 21 Jun 2012 22:02:27 GMT

Hi Lawrence...

Most companies I've ever worked with tend to avoid merges for the purpose you're describing. Having said that, I suppose there's nothing wrong with using a merge to represent a subassembly. But there are numerous other ways to achieve the same effect and each have their upside and downside.

You've got a sticky problem. Let's say you have a regular part (a square block) that belongs to a regular assembly. At the start of the day, you check out the part and add a hole to it. You check in the part and erase everything from memory. If you were to open the assembly, the block part would have the hole in it automatically... and without regeneration.

Now let's say you need to create a merge part of two blocks side by side (as if they were a small assembly). This is similar to your situation. The blocks are again square with NO HOLE. There are a couple of ways to do this, but let's say you open a new part and select Insert->Shared Geometry->Merge/Inheritance. You pull in two of the block parts to use as your merge references. After pulling these in, you've now merged them into a single part... and you check in the new merged part. Next, you check out the block part, add a hole, and check it in. You erase your memory. When you open the merged part, the blocks do NOT show the hole.

This is essentially your problem. And... I haven't found a good way around it. Because the blocks in the assembly are the ACTUAL blocks, they update. Because the blocks in the merged part are only references, they do not automatically update. I'm sure this is by design but it's causing you grief in this situation.

Perhaps if you took the assembly and created an Envelope part from it, that might update as the individual parts did. You can use an Envelope part to represent an assembly in an upper level assembly. The intent is that you use the envelope to simply this "upper level assembly" but you could preserve all the details of the individual parts if you wished. I'm not very sure even this will work.

Here's what I know... or at least what I think I know: Shrinkwraps, Copy Geometry, Inheritance/Merge, and Component Operations->Merge in assembly mode are all going to have the same problems. External copy geoms/shrinks/merges will also suffer the same fate. You could export the assembly as a STEP and bring it in as a part to create a "merge" but this is obviously going to prevent automatic updates. Envelopes may work... but I suspect not.

So then... maybe we need to back up. WHY do you need a singular part to take the place of two separate pieces? Is this an "inseparable assembly" and your company just prefers to see it as a single part because of this? If you're truly just "representing" the assembly as a part in upper level assemblies, then this is truly a job for simplified reps with an envelope part being substituted in for the subassembly.

I feel like maybe I'm missing something but I also feel like I don't quite have enough information on the situation to really pin down the problem. Can you give me a bit more guidance and I'll rummage through the trash bin of useless information (also known as "my brain") to see what I can come up with.

Thanks!

-Brian

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Fri, 22 Jun 2012 23:20:04 GMT

Thank you for the detailed & helpful reply. You touch on some good points that I am discussing with a coworker. I will try to reply back early next week and give you some of the missing details and perhaps why we started using a singular part in the hopes that we can solve these Regen issues!

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Wed, 27 Jun 2012 18:25:54 GMT

Sorry to get back to you so late.

You are correct, we are showing this sub-asm as a part because it is inseparable. After the subasm goes to get brazed it comes back as a single part. Being only one part, this simplifies the top level asm dwg display where the x-section and hatching are shown, while appearing more accurate (no brazing rings, and previous components are joined into one part). The user only needs to enter one hatching for the entire part without having to go thru each part. Is this an option with simplified reps to tell the system to behave like one part?

So as far as I can tell, a summary of the main advantages of using a separate merged part is:
1) Not showing the brazing rings,
2) Clear/easy X-hatching
3) 1 part/joined display.

The sub-asms we deal with are very simple, usually involving only a few parts, in addition to a couple of brazing rings. The only component difference between the asm and merged part are the additional brazing rings that fuse the parts together.

I have never used envelope parts before and have been trying to look into them since you suggested it as a possibility. As I understand it envelope parts are to simplify complex asms. Should envelope parts be used when the subasm is already very simple? If so, how would you use them? I tried playing around with them and I get an error saying that an envelope part can only be used in the asm it was created in so I don’t know how to use the part once it is created. Also, what advantage does using an envelope part offer beyond just creating simplified reps in the sub-asm and then using them in the top level asm? Do you know the difference between envelope parts and envelope assemblies?

Once I started thinking about this, if we do not use a separate part to represent the joined sub-asm then we could use FTs to show the sub-asm without the brazing rings, or for that matter even use the normal sub-asm and then make it flexible in the top level asm in order to suppress the undesired brazing rings.

I don’t understand why the merged/inheritance parts not updating automatically would be by design. When creating the feature the user has the option that it be dependent or independent. Independent would obviously be for a case where you don’t want the child part to update with the parent part, but choosing dependent seems like it should be the way to tell ProE that we do want the components to automatically update. Maybe this is more difficult than it sounds to implement but it seems like this should be the goal in order to make these features more useful (even if only by giving this as an option). Have I missed something? You are certainly right though that whether by design or by poor design, it certainly is giving us a lot of grief .

Thanks again for your help with trying to find a better solution to these regen errors!

Re: Regeneration errors in Merged Parts/Assemblies?

BrianMartin — Wed, 27 Jun 2012 20:03:17 GMT

Thanks for the additional clarification. I completely understand what you're doing now. I wish I could say that this has opened up some great technique to get you to where you need to be. Unfortunately, I don't think it has.

First, let's talk about envelopes. Envelopes are used to simplify complex assemblies. There are other tools which can also do this. For instance, you could use an Interchange to do this. There are some unique options available when creating the envelope that are pretty handy. For instance, you can quickly create a 100% accurate copy of all solid surfaces selected for the envelope. You could do this manually with a Shrinkwrap or a Copy Geometry feature but the process is lengthy. The envelope provides an automated method to do this.

You asked several questions about envelopes so let me try to address them. I should mention I don't often use envelopes but I have had occasion to use them. Like any other tool, they have their place.

You asked if envelopes should be used when the subassembly is already simple? Probably not. If you had a large, complex model, envelopes would be more useful. The more I think about it, I don't like the idea of using an envelope to represent an inseparable assembly. The envelope does satisfy the requirement of simplifying the cross-section. It looks and behaves as an individual part so it has that benefit... but I think otherwise it falls short.

You CAN use an envelope part in an assembly other than the one in which it was created. In the image below, notice the envelope feature in the assembly is called ENVLP001 but the part is actually called Envelope_Part.prt. You can use Envelope_Part.prt in other assemblies.

The main benefit to the envelope is just as a vehicle for simplification. At least, that's the only way I've ever used them. There are several side topics we could discuss regarding envelopes... such as the ability to load a standalone model and use it to represent an assembly. You could model a single part that resembles the fully brazed assembly and use that part as an envelope representation of the assembly. But this would be duplicating work and I think it's taking us down the wrong path.

Your idea about using flexible components or family tables would likely solve your problem with components not updating. But then you've added the overhead of managing the table and/or the flexible features. It may be trading one headache for another. It's worth taking a closer look. Maybe you'll get lucky and strike the right balance of automatic update and ease-of-use.

To discuss the merge/inheritance parts not updating, maybe I can shed some light on things. Yes, you have the dependend and independent switch. This does work... but it only works when the references used to create the merge/inheritance are in memory.

For instance, let say I have an assembly. I create a new part (call it part "X") inside the assembly. Next, I activate part "X" in the assembly and copy features from various assembly items using copy geometry, shrinkwrap, inheritance, etc. Now I have an assembly containing part "X" and part "X" itself contains copied surfaces from the assembly.

If I open up part "X" all by itself, I see my copied surfaces. If I switch back to the assembly, move objects around, and regenerate, part "X" updates. Finally, let's say I check the whole mess into Commonspace. The other parts making up my assembly could be changed and altered... but part "X" will not update. I could open the assembly itself and make changes... but if I do not check out part "X" and allow it to regenerate, no updates will take place even if I have dependencies turned on. I MUST regenerate that assembly before part"X" will update and part "X" must be checked out.

If the references in part "X" are set as independent, they won't update at all, even if I open the assembly and regenerate. All that dependency switch does is allow part "X" to update if it's checked out, if it's parent object (the assembly) is in session, and if there's some change to be picked up during the regeneration.

The only way you seem to get an object that appears in it's modified state without regeneration upon it's retrieval is the modifed part or assembly itself (from the example in the previous email).

After doing some additional research, I only had two ideas that seemed to give you what you needed -but each has a pitfall. First, do the merge. This gives you a single part... yet also gives you individual components of an assembly as they exist in real life before the brazing. The venerable Leo Greene has a nice video showing his technique for producing a new merge part from several pieces. He also demonstrates incorporating welds and other features. Here's a link to the video: http://www.e-cognition.net/pages/MergeModel.html

The second idea is less desirable. You can choose to create your inseparable parts ... as PARTS... but using merges. Let's say you have a plate with some PEMs or pressed-in inserts. The inserts are separate Pro/E models as is the plate. Obviously, they're all parts. It's possible to make a PART look as though it were an assembly using merges.

For this example, you'd open the plate part. While still in part mode, you'd select Insert->Shared Geometry->Merge/Inheritance. From the dashboard, use the File Open icon to open one of the insert models. You have to constrain the insert model just as if you were assembling it into a real assembly. Once this is done, hit the check mark to complete the feature. The ENTIRE insert will be copied into the plate part as an external.

It's hard to believe but you can make a part that looks like an assembly using this technique. (see below). The downside here, is that you do not have an automatic BOM for your inseparable assembly. It's just a part... so it functions like a part (no BOM). For this technique to work, you'd have to fake in a BOM (which I never like to do). If you make changes to the base part (the plate in this example), it does update without having to regenerate. If you make updated to components merged in, you still need to regenerate to pick up the modifications.

I'm not sure we have any other options that could work. Take a look at the options and please write back to let me know what course you decide to pursue.

Thanks!

-Brian

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Mon, 16 Jul 2012 20:01:19 GMT

Thanks again for the great reply and ideas. I too wish there were a more robust solution.

Since there is not a straight forward solution I think for now we are not going to change anything and keep using the Merged prt. I will play more with the envelope parts to see if I can get that to work for us, but we have a lot going on. We just took the plunge into using PDMLink 9.1 for document management to include non-cad files as well and that is proving to be rather interesting...scratch that, painful. We are also looking to upgrade ProE to Creo and PDMLink to 10.1 so I think we have a thorny enough road ahead of us to temporarily keep us away from delving and and fixing these annoying but smaller problems.

Thanks again for you ideas and help. I REALLY appreciate your help! I will try to get back to leave another reply in this forum if I figure out a better solution.

Lawrence

Re: Regeneration errors in Merged Parts/Assemblies?

linda — Wed, 18 Jul 2012 14:31:27 GMT

I do not think it is a good idea to merge a sub-assembly to an assembly.

it add an external reference which often causes problems,and it will cause problem when uses repeat region for bom with part level . you need many manual jobs to make the bom correct.

but if you do insist to use the merged part to represent a sub-assembly, you can use the option "retrieve_merge_ref_parts" to yes, which will retrieve all referenced parts to the session, and when you open the part with external reference and the status icon at the bottom of the screeen is yellow, you need to regenerate the part,if green, not referenced part changed, no need to regenerate.(creo 1.0)

hope it will help a little bit.

linda

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Wed, 18 Jul 2012 14:46:21 GMT

Thanks for the response Linda. I am not sure what you mean by merge a sub-asm to an asm. This does not sound like what we are doing. We merge a few individual components in a subasm to each other, thus forming a single merged component. We then use that merged component in a subasm. The top level BOM needs no manual manipulation because we want it to show as a single component. As for the subasm dwg, we do not show the merged part on the dwg, so the auto BOM reads the components from the subasm and NOT the merged component. Do you still advise against this? If so can you ellaborate more?

If I understand you correctly you are saying that if my merged asm is showing that it needs regeneration, I can

1) change the config option "retrieve_merge_ref_parts" to yes,

2) manually regenerate the merged part and the regeneration will not effect the merged part...(kind of like using the flexibility feature in asms are not supposed to affect the components they are changing)?

I will test this on the next merged part that I have to fix. I really hope this works and I really hope it works on wf 4.0. If it doesn't work in WF4 I will also test it in Creo 2.0 which we are using experimentally right now. Either way if this works it sounds like an ideal solution and would cut out tons of wasted time. I will let you know. Thanks!

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Wed, 18 Jul 2012 23:04:05 GMT

I just tested changing the setting "retrieve_merge_ref_parts" to yes in both WF4 and in Creo2. It didn't seem to change anything as the merged part still says that it needs to be regenerated when one of its reference components changes.

I even tried locally changing the the merged part then manually regenerated, saved, erased from session, opened one of the parts used in the merged SA, force a regeneration, saved, erased it from memory and then opened up the merged part but again the merged part still needed manual regeneration.

Am I doing something wrong here? Is there another way to get that config option to make it so our Merged parts do not need manual regeneration when one of the components in the merge get regernerated or changed? Any other ideas?

Thanks again for your help and PLEASE let me know if you have any other ideas on how to do this.

Re: Regeneration errors in Merged Parts/Assemblies?

linda — Thu, 19 Jul 2012 18:20:00 GMT

i mean to change a sub-assembly to a single part, as your case.

this option set to "no", you open the top assembly with a meged part from a subassembly, the regenration status is green even with your merged parts out of date.even when you regenerate the top assembly, the merged part is not updated if the referenced part is not in the session.

with this option set to "yes", when you open the top assembly, the referenced parts are loaded to the session and update of your merged part becomes available. and more important, the regeneration staute is yellow, which tell you in advance that your model is out of date, you need to regenerate the models. not later when you finish lot of jobs, you realize that the some models are out of date.

only with the option set to "yes", and when the status is green, you need not to regenerate the top allsembly, otherwise, you risk working on a outdate model if you do not generate the model, or even you regenrate the model whicl the referenced part is not in the session.

for the bom of top level, you can have correct data only if your merged part is updated, if it is outdate, you get wrong info, general wrong mass info.

for bom of sub-assembly, hide the merged part in the drawing do not prevent its mass be added to the subassembly, you have to suppress it from the assemnly to get the corrrect mass info in the bom.

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Fri, 20 Jul 2012 16:40:27 GMT

I think I understand what you mean now. So neither YES nor NO actually regenerates the merged model and components.

If I understand the config option correctly, Setting "retrieve_merge_ref_parts" to:

--Yes--Does not regenerate components but NOTIFIES user that they are NOT regenerated

-- No --Does not regenerate components and PRETENDS everything is ok by using old components while NOT notifing user that the components are NOT the latest and NOT regenerated.

I see why you say to stay away from merged components. Unfortunately we are pretty fortified in them and if we do move away from them we need to have a good option to make it worth while since it looks like alternative methods would all require fixing our top level asms and all their constraints...ouch!

I hope PTC adds an option to retreive and regenerate (even if a warning message was shown) so there is the option for merged parts to behave more like the asms they represent. Alternatively it would be nice to have another method to get similar results, which seems to be a mixture of part and asm. It would be awesome to migrate from WF4 --> Creo(3?) and to see this fixed!

Thanks again for your help. If you or anyone else has more ideas, I am all eyes and ears!

Re: Regeneration errors in Merged Parts/Assemblies?

GaryBarclay — Tue, 31 Jul 2012 09:10:21 GMT

Stick with the 'Envelope Theory' it is the way to go ! Use the option 'all solid surface shrinkwrap'. After you have selected the components you want to be in envelope (everything excluding the brazing rings) go into options tab

and solidify everything into one part (you can set the dependency on or off here). You can use this envelope in your drawing and all the sections will work well. Check it out in LMS

Re: Regeneration errors in Merged Parts/Assemblies?

SylvainA. — Tue, 31 Jul 2012 11:06:33 GMT

Hello Lawrence,

I just tried and for me the solution is to make External merges (Insert>Shared data>Merge/Inheritance) in your part (instead of merging parts in an assembly) and to set the option retrieve_merge_ref_parts to yes.

When you set it to yes, Pro/E retrieves merged components and update the geometry of your "sub assembly" part.

If set to no, Pro/E says "Reference model XXXXXXX for feat (id 383) in part YYYYY not in session.

WARNING: External ref. for feature/component not found, using old placement"

I'm sure this will work for you

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Tue, 31 Jul 2012 19:20:13 GMT

Thanks for the suggestion Gary. I just tried using the envelope part in Creo 2.0 using a couple of different ways and I keep having the same regen issues as when I use the merged parts.

I tried creating part and shrinkwrap envelope part while in the asm
I tried creating a shrinkwrap in the part and then using that part in the asm.
I tried creating a envelope part using the view manager and envelope manager while in asm mode

All 3 trials required a regen to the shrinkwrap part once a component. Have I done something wrong? Is there a way to do this so that the assembly part does not require a regen when a component changes?

I would really like to upload a sample asm in ProE/Crep but don't see a place to add attachments. Any suggestions? If you or anyone else would like to give it a shot I can send a sample file to you.

Thank you!

Re: Regeneration errors in Merged Parts/Assemblies?

BrianMartin — Tue, 31 Jul 2012 19:26:08 GMT

Hi Lawrence...

As far as I can see, nothing has changed and you're at the same standstill you were previously. I think you have a firm grasp on the limitations you're working with. I wrote originally that envelopes were a way to go but that I didn't like them upon closer inspection.

After some other users' suggestions, I went back and read what I wrote. I have to stand by what I said previously. I see what you're trying to do and why... I just don't know that there's a better way to get an automatically updating merge model than what we're already tried.

If you'd carre to submit this as an idea or enhancement request, I'll be more than happy to bring it up at the next Technical Committee meeting and see what the PTC developers think about adding this feature. What you need is a Merge with a dynamic update option. Sylvain's idea is close but I still think this doesn't get you where you need to be.

Thanks!

-Brian

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Tue, 31 Jul 2012 21:31:13 GMT

Thanks for the suggestion Sylvain. I tried what you described using an External merge instead of merging parts in the asm and set the config option to yes.

If I understand this correct I would have 2 separate models:

a PART with the merged features created directly from the original components (as opposed to connected to the parts in the asm)
The ASSEMBLY which would be displayed on the dwg only so that we can display and call out all the components in the BOM

It seems that if we used this method, we would have to create the part twice since the part would not be dependent on the asm.

Also, when I tried this, I still had to manually update the external merged part in order for it to get the latest components, even with teh confic option set to yes.

Is this what you were describing, or is there another way to do this?

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Tue, 31 Jul 2012 23:22:27 GMT

Hi Brian,

Yes, there have been a lot of good suggestions from diferent posters, starting with yours. I understand several of the available features in ProE/Creo much better now yet unless I missed something from the methods that others have outlined, this appears to be a limitation in the product (as you said) and the only good solution appears to be in the hands of ptc creating a dynamic update option, or equivalent.

Actually, some time after your last post I submitted this as a Call to PTC. I debated as to whether it should be considered as a Problem or as an Enhancement Request; I finally settled on letting them decide; if they don't consider this a problem then they can swith it to an enhancement request.

From our perspective it is an error for the following reasons

The "dependent" option only works by manually updating,
The config option retrieve_merge_ref_parts set to yes (ie. Automatic reference parts retrival for dependent Merge) doesn't actually retrieve merge reference parts but only notifies that they are out of date. (after all the config option is doesn't say notify_merge_ref_parts).

Incidently the config option Regenerate_read_only_objects set to yes also seems like it would regenerate read only objects (instead I think it only notifies that it needs regenerating and allows the correct display of components that use the flexible feature).

That being said, if anyone knows of a solution, or if I misunderstood a solution from another poster, I am still all eyes and ears! Please help me understand. Thank you!

Otherwise, I hope PTC is able to make this improvement/fix.

Re: Regeneration errors in Merged Parts/Assemblies?

SylvainA. — Wed, 01 Aug 2012 06:23:07 GMT

Lawrence Scheeler wrote:
It seems that if we used this method, we would have to create the part twice since the part would not be dependent on the asm.

Why twice?

As you said, you need 1 assembly for displaying the BOM and 1 part which will be the merged S/A (like, I suppose, you are currently doing).

I agree that the assembly process has to be done twice, once in the assembly and once in the S/A part.

In the part, each componenet of your assembly will be added as an external merge.

In the picture below you can see how it looks like when you merge a component in a part with an external merge.

It is like adding a component in an assembly but you are still in a single part.

I tried it yesterday and it worked fine for me.

I modified a merged component, then saved it in my workspace, closed Pro/E, re-opened Pro/E, opened the S/A part, regenerated it and the merged geometry was updated...

Re: Regeneration errors in Merged Parts/Assemblies?

LawrenceS — Wed, 01 Aug 2012 16:30:11 GMT

Thanks for the clarification Sylvain. I think your last post helped me understand that what we are trying to do is a little different.

Sylvain Arjona wrote:
opened the S/A part, regenerated it and the merged geometry was updated...

These last two steps (opening and regenerating) are exactly what we are trying to avoid. We want the SA part to update without having to manually open and regenerate it. We just want it to show the latest, just like an asm always shows the latest.

It doesn't sound like a big deal untill you have a component that may be used in up to 100 other sub-asms combined with a data management system which requires you to demote, checkout, checkin, and repromote just in order to regenerate a prt or asm...and have to do this for possibly all 100 S/A parts that the component is used in. This is when it become a tedious, time consuming, royal pain in the neck. So the external merge that you suggested will bring with it the same issue we are having, plus getting people to asm the components twice (and make sure both are always updated.). Does this make sense?

Thanks again for the suggestion, however it looks like the external merge doesn't solve the problem. Any other ideas?

Re: Regeneration errors in Merged Parts/Assemblies?

SylvainA. — Thu, 02 Aug 2012 06:32:11 GMT

OK, now I understand.

Lawrence Scheeler wrote:
..... until working on the top level asm that won't regenerate because of the un-regenerated merged part....

An external merge allows the "S/A part" to be regenerated during top level asm regeneration, which is not the case with a part merged in a sub-assembly.

I thought that was your main issue