I Hate Drawings!!! 12

[Bester2]

I need to rant. Why do we have drawings in the 3-D world? I am so tired of arguing about the line thickness/ the font size/ the angle of the leader line or all of the other BS that goes along with creating drawings. Then you issue a fabrication to someone and they inevitably call me back asking how does this thing go together. Then I send a packaged assembly to them and there is no more questions. All they needed in the first place was the model with the associated material. Weld callouts can be called out as annotations, for that matter they should just be physically modeled. And when it comes to assembly, the model is the easiest way to show how things come together. Today you can even create video clips that can be animated with notes to show how things come together. I worked for a company once where I heard that an entire division was designed paperless. In order to do this they made all of their suppliers run the same CAD package. This allowed them to created annotations in the models along with associated views to make the parts. Do places like this really exist? If so dose anyone else see this becoming this way in the future. Dose anyone else agree with me that drawings are a waste of time? Does anyone think that their mechanical task is better served in a two dimensional world? Am I doomed to suffer in a world of arbitrary existence??

One extremely frustrated engineer that is probably looking for a new career!!!


SW 2007 SP 5.0

 

[ewh]

Depending on the software, all of that additional design info and intent that needs to be conveyed can be conveyed in the actual model. All of it.
We will have drawings for quite awhile though, as many ancillary functions will be slow to adopt a "digital" mindset, and full integration is still quite expensive, but to deny the possibility is to stick your head in the sand.


When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[KENAT]

"Depending on the software" and there in lies one of the major problems - at present.

KENAT, probably the least qualified checker you'll ever meet...

 

[ajack1]

I am with ewh on this one; the technology is there and has been in place for many years now.

As to how many companies take it on board, well I guess only time will tell. I am sure this debate has been discussed with regard to going from the drawing board to 2D CAD or from conventional machine to a CNC machine, how many companies are still in business with a draughtsman at a board and a machinist winding handles?

As with anything you need the other party to be able to “understand” what you are telling them, a phone or an email account are also useless if the person you are trying to contact does not have them or know how to use them, but I am sure most of us would be lost without them and would not use a company that did not have them.

In five or ten years time will saying sorry we need drawings be the same as saying sorry I don’t have email can you post it to me please?

 

[drawoh]

ajack1,

I do not see 2D drawings disappearing that fast. For everyone to get onto 3D, we need the following...

[ol]
[li]Everyone who requires the information has high speed computers with good video cards, capable of managing every bell and whistle that gets added to the 3D[ ]software. By everyone, I mean engineers, designers, drafters, technical writers, inspectors, fabricators, shipping clerks, etc. [/li]
[li]These machines are kept up to date, systematically. [/li]
[li]The aforesaid high-end CAD[ ]machines are portable to where people need to do design reviews, inspection, or whatever. Any networking requirements are met somehow, without security problems. We can drop these high end machines, and pour coffee and crankcase oil on them. [/li]
[li]There is a standard 3D[ ]CAD format, supported by all CAD[ ]vendors, that provides full design capability.[/li]
[li]Blue highlighter capability, as well as redline capability. This is the minimum required for design checking and inspection. It is not obvious to me how you are going to do this in[ ]3D. I am not aware of any 2D[ ]programs that support highlighting.[/li]
[li]Probably, I have missed something.[/li]
[/ol]

Mechanical 3D[ ]software runs at the limit of what most CAD[ ]stations are able to manage. My experience is that mechanical[ ]CAD has always done this, and I do not think things will change in the near future. When the machines get faster, everyone will use accurately modeled screws and springs, and we will systematically use photo-realistic surfaces.

You mention how email has taken over from snail mail. There is not much downside to email. It does not create anywhere near the logistical issue 3D[ ]CAD does.

JHG

 

[KENAT]

drawoh - you've covered many of the points I see as stumbling blocks for many places at the moment and for the foreseeable future.

We are starting to use the model as the primary part definition for some castings and moldings but, due to cad interoperability issues we still have a minimal 2D drawing with the notes and other pertinent information that can't be readily contained in the current dumb model formats that we & our vendors can work with.

It seems a lot of the 3D advocates/current users are sectors with companies that can effectively 'force' their supply train to use certain CAD formats, however for many small and medium players this isn't the case.

Once a true standard lightweight model format that can contain PMI is established a lot of these issues will be reduced but I'm not sure when that will be. UGS had me convinced that JT might be the way forward at one point but now I'm not so sure.


KENAT, probably the least qualified checker you'll ever meet...

 

[ewh]

drawoh,
I think you may be a little behind the times.

1. Everyone who requires the information DOES NOT require high end computers capable of managing every bell and whistle that gets added to the 3D software. All they need is a good connection with the company intranet and acess to a good PDM package. They do not have to load the actual CAD model, only a light weight representation of it.

2. Technology has less to do with this than company discipline, with which this should not present a great problem.

3. While still more expensive than the typical laptop, hardened laptops have been manufactured for field work for years now, and encrypted networks have also been around for awhile.

4.This is the one point where I tend agree with you, but it would not surprise me to see inter-CAD compatibility as to MBD before very long. The PDM software will be more of a driver than the CAD software.

5. Already exists; has at least as long as electronic approval has.

"Mechanical 3D software runs at the limit of what most CAD stations are able to manage."

This is very true, but what you have overlooked is that the only person requiring such hardware is the one creating or editing the model file, not the end users, and that processing power is still ever increasing (I forget the name of that constant).

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[KENAT]

PDM package - that's one of the hurdles for many places.

While I'm convinced of the benefits of one some management aren't, and to do it well again you want interoperability with your vendors, as I understand it.

KENAT, probably the least qualified checker you'll ever meet...

 

[ewh]

Your preachin' to the choir with me, KENAT. While local management here is sold on the idea, corporate still needs some nudging. As if countless duplicate files and repeatedly loosing primary released data wasn't enough. Add to that the fact that several of our biggest customers use the software we are considering. I think that corporate IT just doesn't want to be told what it needs to do unless it comes up with it first.

While interoperability with your vendors would definitely be a good thing, you can still use whatever model files they deliver to you. Compatability is the biggest hurdle as I see it, and as drawoh mentions in his 4th point.

Again though, that does not negate the fact that a paperless system is possible to accomplish NOW. You just have to overcome a couple of remaining high hurdles. With time, I am convinced that it will also become practical, without having to be a GM or Boeing.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[drawoh]

ewh,

Your are refering to Moore's law. Drawoh's law is that I want more stuff out of my CAD system. I want to model screw threads, and I want to install my screws by screwing. I want photorealistic surfaces that account for anodizing, chemical films, zinc plating, translucent materials, and fabrication process. I want everything to work in real time.

Once upon a time, I worked on complex assemblies on 2D[ ]AutoCAD on a Sparc[ ]LX I think. Something like that. I would turn off the autoupdate and mess with all sorts of blocks. Then, I would force an update and go off for a coffee.

All of this would work instantly on the computer I have now, but I am not on 2D[ ]AutoCAD anymore. I am on SolidWorks. I spend a lot of time on Eng-tips because I am waiting for stuff to load or update.

You add power to my computer, and I will find a way to (ab)use it.

JHG

 

[ewh]

drawoh,

Moore's law... thanks!

I've been using UG for over 20 years, and Applicon and ComputerVision before that. We are light years ahead of where we were way back when, and I can only think that such progress will continue.

I remember regenerating a shaded view, being able to take a 30 minute break and returning to my desk before it was finished. It was handy when cpu usage was checked to see how hard you were working ;-). It was also easy to bring the mainframe down, bringing many others to a standstill.

We don't have those problems with UG today. I now find myself getting frustrated if a command takes more than a couple of seconds to execute. It runs well on high end laptop computers, and even runs well on Apple computers, using an emulator. It can create lightweight models which can be fully interrogated and marked up to your hearts desire. The software is becoming easier to use than ever. Screw threads, though, still bring a lot of baggage, but it will automatically create them if you want them.
I'm not a salesman, just a user, and I look forward to the ever-increasing improvements that I am seeing.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[KENAT]

Moores law has hit a bit of a snag though, to get around this they introduced the multi core processors.

Trouble is many aspects of most CAD programs can't take advantage of multiple cores/parallel processing so this limits increases in performance at this time.

KENAT, probably the least qualified checker you'll ever meet...

 

[drawoh]

KENAT,

The problem is that Moore's law really ought to be called Moore's observation. The technology has to plateau sometime. At some point in time, you have processor tracks one atom wide, with electronics travelling on them at the speed of light.

I only hope this happens after they implement my wish list, above. Also, I have to be able to play rock videos off the internet while my CAD software does everything at high speed.

JHG

 

[KENAT]

You don't ask for much do you;-)

KENAT, probably the least qualified checker you'll ever meet...

 

[DesignerGuy16]

Problem is you can't really "lock out" models in the same way you can encrypt PDF's. That's why they're the prefered release file type. Because anyone can view it with a free reader, and almost any computer under the sun can run the software.

Take a case like my current workplace. We have a pretty decent sized machine shop. For what you're saying, we'd have to get another 20 seats of SW, and train the machinists to fumble through the software, with computers that can handle them. Now we aren't even on subscription with the 6 seats we actually use on a daily basis, how do you think we'd be able to afford setting up the whole plant with it? Yes I know there's edrawings, but it doesn't have the same capabilities with measurements, etc. Unless you're making an edrawing of the drawing. Then you're still back at square 1 of having to make a drawing that depicts the correct design intent.

Or, we can keep making drawings, and having our CNC programmer take our models and drawings and create his programs. This still seems like the most reasonable route to me.

It all sounds great in theory, but unless you're doing very complex surfaces it doesn't really pay dividends imo. CAD has come a very very long way towards putting the design intent in the model and hitting a "draw it for me" button to basically copy all of the design intent from the model to the drawing so there's minimal detailing time. I forsee this being more of the direction as machinists and QC will always need some form of drawings to check to.

James Spisich
Design Engineer, CSWP

 

[ewh]

There's that word "always" again, and I must respectfully disagree.

It is not the responsibility of departments other than engineering to make ANY model changes to released files, so it can not be justified to require the use of the software that created these files for them to complete their tasks.
Our machinists can take our released model and create the CNC program. They can not change this model (it IS locked out as far as they are concerned), but their software does allow them to read all of the model based annotations and properties. They do not need a full blown seat of the software, just the CNC software. They will create a new CNC file which will contain and refer to the parent model file.
Our QA department can use that same model to create a file with the inspection annotations and criteria, again without being able to modify the actual model. Again, they do not need a full blown seat. They only need to have the ability of annotation creation in a file which contains the parent file.
If and when they find problems with the parent file, it is the responsibility of engineering to rectify the situation, not their own.
This ability is not theory, but reality. The largest problem, as has been noted before, is to have everyone using concurrent software. If everything you need is made in-house (with the exception of raw materials), and you've got the $$, then this is very achievable today.

Those "draw it for me" commands are as dangerous as allowing release of unchecked (or peer checked) drawings made by untrained (in drawing creation) engineers. They don't know what a technically correct drawing is much of the time, and automating the process is just going to exaberate the situation. But that is a topic which has been well covered in these fora.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[Bester2]

On the 45th anniversary I would like to pay homage to the late MLK.

I have a dream where computers talk to computers and all human error is taken out of the equation.

I have a dream where the geometrical electronic representation is beamed via wireless communication to an electronic computer controlled machine to make various machined parts.

I have a dream where computer controlled measuring machines can take the same geometrical electronic representation and created programs to check critical dimensions on the machined part that was just created by the computer controlled machine.

Oh I have a crazy dream where the entire definition of a machined parts is defined by the creating engineering group and additional human errors are all but eliminated.

45 years ago Dr MLK spoke of his dream and even though I believe that we have made strides towards his dream we are not 100 percent there yet. I have a feeling I could reread this great work that I have just written and still see that while we may have made great strides towards seeing a world like this the reality is that this will probably be just what I have suggested it was, a dream.

Please no nasty responses this was all very tongue and cheek, it was written to make you laugh.


SW 2007 SP 5.0

 

[DesignerGuy16]

I've dealt with enough engineers who were terrible at drawing creation. Typically their modeling practices were FAR worse then that. At least if they're terrible modelers and decent at drawing layouts it's not SO bad.

Just doesn't give me warm fuzzy feelings is all. Depending on your industry and what you're making I don't think you'll ever get around it.

James Spisich
Design Engineer, CSWP

 

[ewh]

But, for any future editing and to implement MBD, terrible modeling IS bad, and not at all cost effective.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[drawoh]

Bester2,

In a world of 3D[ ]modeling, you do not eliminate human error. You just change the human errors into different human errors. 2D[ ]drafting is a language, with definitions and syntax. 3D[ ]modeling will have to be a language with definitions and syntax, so that you can write purchase orders and tell people to fabricate widgets as per file 123456.sldprt. The designers, fabricators and inspectors must be on the same page as it were.

The 2D[ ]drawings generated from 3D[ ]models insulate us from some very bad modeling practise.

JHG

 

[MechEng2005]

Another issue I see is with providing drawings to customers. Where I work, we typically provide only assembly drawings, without details of the parts. Basically, we include drawings to show how the machine works and dimensions/information needed to use and maintain the machine. Anything that would only be necessary to duplicate the machine is NOT given to the customer. One reason is that we do not want this information available outside the company, where it could be used by competitors to reverse engineer the project.

If the only information is 3D models, with measuring capability, then providing the model to the customer is giving EVERYTHING. Sure, people may be able to scale from a typical drawing (although typically the printer is set to "fit to page" so you end up with some weird scale), but it would be difficult and chances are that some details would not be available from just assembly drawings.

I suppose a "dummy" assembly could be created. I know that some manufacturers of engines and things provide a shell of the acual engine. The model doesn't have the pistons/bores or internal workings. However, creating a "dummy" assembly would cost some of the time saved by not creating drawings, and (at least for components without some form of housing) wouldn't be very successful.

I don't see overcoming this obstacle without drawings. And I think that would be a major issue for company presidents and corporate managers...


-- MechEng2005