Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Validating 3D models from 2D drawings 4

Status
Not open for further replies.

shaneca6

Mechanical
Sep 27, 2010
12
0
0
US
We recently switched from Autocad to Solidworks. We have a sister company in India that is going to convert the majority of our 2D drawings to 3D models as well as recreate the drawings. The problem I am having is finding relevant information on how to validate all of the data coming to us from India. Does anyone have knowledge in this area? I could use all of the help I can get! *girl smiles.
 
Replies continue below

Recommended for you

To my experience the only way to be sure is to manufacture a set of products using the new 3D and 2D drawings and check the part according to the old and the new drawings. Only when the parts pass this thorough QA then you can declare the new manufacturing drawing a "Approved for production". I have seen hundred thousands of US dollar gone to the drain when this process was not followed. As more the parts are complicated the need for this process is a must.

Just for example: let assume a pneumatic or hydraulic manifold with many drill passes for high and low pressure fluids. One mistake in the depth of a drill can connect high pressure line to low pressure line or return line and you can throw away hundred thousand of US dollar. This is not a dream this was actually happened in a company I use to work for when shortcut was made in the process of drawing replacement.
 
To make this process easier the new 2D manufacturing drawing has to be exactly the same. Meaning that the views and dimensions locations and orientation have to be identical as possible.
 
Thanks for the post. I am actually coming into the project a few months after the initial launch. I have been trying to develop a check list of sorts for the checkers to go by so that the validation is consistent and thorough. I am afraid we went about this migration the wrong way so now it is a matter of getting back on track. I am going to scour the web for all of the tips I can find and hope to put something together for the team to review by the end of the week!
 
I agree with isaelkk, you have to be able to do an "apples to apples" comparison between the old and new drawings, as well as compare new parts with legacy data. I'm not sure of the scale of your project, but if possible, it's always better to handle a migration like this on a product by product basis until some confidence in the process can be established.

I was involved with a switch from AutoCAD to SolidWorks at a previous company. After discussing the details and pitfalls of the move, it was decided that all new products would use SolidWorks, while all existing products stayed in AutoCAD until an engineering change was needed. During a change the conversion to SolidWorks was made and verified.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?
 
I do the same as MadMango.. New products get completed in (Inventor-3D). Old products only move out of the Autocad 2D world when an engineering change,etc... is needed.

If you can't trust that India will recreate the drawings 100% then I would require a 100% check of all drawings to verify correct dimension/tolerances/notes/materials,etc...

Like Israelkk said..1 simple mistake and you can literally throw all your money saved by outsourcing the work right down the drain with the bad parts.
 
We currently have a little over 12K 2D drawings that need to be converted to 3D which is why we contracted our sister company to do this work for us. In the middle of converting the 2D drawings, we are also transitioning from Innocielo (AMT) to ePDM. We have had many meetings in recent months to determine how we handle the previous revisions of drawings. I have suggested archiving them and leave them as they are. Others in my organization want to correct the revision levels to make the data clean and usable. I believe this method is going to cause more problems than solutions.
 
You're sending technical data (drawings) to India. They are taking that data and inputing it into a 3D modeling program, then from those 3D models they are generating new drawings. The new dwgs should be the same as the old ones (unless you're changing your manufacturing standards).

Actually, there may not be any need for new drawings unless some different information is needed that SolidWorks can represent better. The technical data should not change. Don't let the somebody all geeked up on SolidWorks try to claim that it generates a better drawing. It's a piece of paper with lines and words on it. SolidWorks' funtionality to generate new drawings and views might come into play on new parts or changes, but unless those features are needed now, the critical thing is to ensure that the 3D models are correct to your existing designs. The easiest way to do that is to have SolidWorks' outputted drawings exactly match the existing ones.
 
shaneca6,

Remodeling your current 2D AutoCAD drawings in SolidWorks sounds to me like a good way to train your people in-house on SolidWorks. I think you have wasted an opportunity here.

I suggest good, old fashioned drawing checking. Get a blue highlighter and check each drawing view by view and dimension by dimension. An organized person should be able to do this quickly.

If you were to do this in-house, your SolidWorks drafter would be able to flag any drawings that looked weird or inconsistent, and you would be able to fix any mistakes you spotted.

Critter.gif
JHG
 
Also be prepared for a phone call from India after a newly created model looks nothing like the 2D print because someone was too lazy to redraw the view and simply changed dim values.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?
 
Drawoh,

With over 12K drawings to be converted as well as 50-75 new models monthly, our guys are getting a ton of training. Even the CNC programming guys are responsible for converting some of the drawings to models so they can apply what they learned in training to our real life applications. These replies are tremendously helpful! I am glad I found this forum.
 
MadMango,

Just in training class, we have found several drawings to either be over dimensioned or under dimensioned. Luckily, we have one of the engineers from India embedded with us for a few months during the early phase so he is helping us determine what issues should be dealt with accordingly. We are seeing minor mistakes being duplicated so deciding what to do with them will be carried across systematically.

This experience is definitely one that I will carry with me for a very long time (not to mention added to my resume...*girl smiles).
 
In summary, you're screwed.

We did something similar at previous employer, moving 2D drawings into Pro E and it was a complete ball of chalk. Got contractors into help with schedule so the learning aspect was minimal. Pro E just couldn't do a lot of stuff the way hand drawn 2D did so changes were forced. Went something like 3X over budget or the like and introduced new errors to the pack, this for something that has been in production for 20+ years.

We also did some selected 'high value' items from 2D into Solid Edge here, but to compound things we also tidied the drawings, because frankly they sucked and couldn't just be sent to any competent shop.

As well as the 'mistakes' you mention, there may be a few things SW just doesn't do like the old 2D package and so you can't get a 100% match.

Assuming you aren't incorporating any design changes, tidying up of drawings or the like then I'd suggest 100% checking by a dedicated, qualified checker, of the new SW print against the old 2D print may be sufficient.

If you aren't changing dimensioning schemes, or making changes, then the need to actually make hardware off of the drawings is arguably reduced, but it's sure nice if you have the budget. I'd prioritize this on the drawings where 'errors' were found.

Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
 
Kenat,

I was very afraid of your first statement. *girl laughs.

I said the same thing. I don't even want to start talking about the errors in our bill of materials. That is a subject in need of a couch, a 2:00 appointment, and a few pale ales...or tequila...!

I felt like leaving 2D as 2D unless changes were required would be the right thing to do but the person that made the decision to convert the entire vault was about 4 steps over my head on another continent! So.....here I am trying to make the best of a bad situation and just get through to the next phase!
 
One problem with 2D-to-3D is that it's so much easier to cheat in 2D. Dimensions that were possible because 2D didn't take all the geometry into account become impossible when attempted in 3D.

Also make sure the folks doing your conversion pay attention to actual dimensions and tolerances. Many are content to just hack-and-scab away making 3D directly from 2D geometry without check whether inout or output is accurate.

Expect lots of questions. If you don't get questions, expect problems.
 
Why are you doing this?

You have perfectly good 2D drawings that you are using or have used in the past to make parts.

You (or your vendors) already have CAM/CNC files for these parts.

What does the conversion from 2D to 3D model and then to 2D print again gain you?

Nothing that I can see.

New parts, new system. Legacy parts, leave well enough alone.

 
She's doing it because
<pejorative>TOP MANAGEMENT</pejorative>
already decided, of course.

No amount of derision is appropriate; there _isn't_ enough.




Mike Halloran
Pembroke Pines, FL, USA
 
Well of course it was some idiot that can't even spell 'CAD' that made the decision. No one with a complete brain would make a decision like that, clearly a lobotomy had been received.

Your idea of updating as needed has some merit, though it can cause its own problems as it ripples up through the pack.

One important thing is to have some rules of modeling, especially regarding the 'fudging' aspect, especially if there is a desire to use the models to help create CAM programs.

Thinking about it some more, getting real parts made from the new drawings/models might not be a bad plan.

If the initial 2D drawings are in any way 'dubious' then it will likely be a real pain.

I've got one 2D drawing from several years back that I was meant to model, redraw & release. I've never been brave enough to ECO it 'cause it's just not clear where some of the dimensions are too, and it's a kind of inseparable assy which is tricky to model. Fortunately, the enthusiasm for that task died back in 07 when I got given the checking role and the bottom fell out of the market for affected products, so no one's asked me about it lately.

Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
 
"but the person that made the decision to convert the entire vault was about 4 steps over my head on another continent! So.....here I am trying to make the best of a bad situation"

And that person will get a slap on the back and a bonus/promotion for "moving the company forward" instead of a kick in the backside for pouring money down the drain. I hate corporate weaseldom, but you might be well served to document dates, meetings, problems, decisions, counter-arguments, etc. in full CYA mode in case someone tries to pin fault on the person overseeing the directive (you) instead of on the person who issued it. It won't work, the management/executive fraternity watches after its own and is generally blind to their stumblings, fully content to lay blame on operations or engineering. You might get lucky and there be one VP or something who moved up from engr or ops and who will understand you're carrying out a flawed policy.
 
tr1ntx,

Luckily we do have said VP...who moved up from Engr or ops and who will understand I'm carrying out a flawed policy. But he has the added pressure from those above him to see this all the way through to the end.

Someone mentioned CAM programs which we do have for all of our parts that are also being re-done from the solid models we are creating. They want everything Engineering does to drive what all of the other departments do. The idea is genius in theory and part of my responsibility is to build workflows aside from the migration.

Whew...I wanna cry! Just saying it in my head made me panic!
 
Status
Not open for further replies.
Back
Top