Good drafting practices, GD &T, and culture shock 12

[mmakm]

Hi all, I'm a lead engineer for a medical device manufacture that has a long history in the automotive and consumer products industry. I have a good background and understanding of GD&T and have been through many coarses. With that being said, the company I'm currently with does not practice GD&T and does not have good drafting standards & many untrained designers (Untrained in good drafting proactices). I'm making an attempt to create standards based on ANSI Y14.5.1 and also instill good drafitng practices in the designers. The questions and issues I have are these:

1. I'm trying to make a good arguement for complete manufacturing drawings because I'm constantly getting resistance to this because we supply 3d cad data and the parts are typically produced from this data. Other than incoming inspection, if the parts fits, the drawing rarely gets looked at. What are other arguements?

2. Does anyone have a good approach to driving GD&T and good drafting practices into a culture?

I was going to add more but if I could get some resolution to these issues, we've taken a huge step. Any input is appreciated, & based on what I've read in these forums we have a great knowledge base.

Thanks,

 

[drawoh]

mmakm,

This model versus drawing discussion is driving me nuts too.

A lot of this depends on what kind of fabrication you do and what kind of tolerances you need to achieve. If you are 100% confident that ALL of your tolerances can be achived by an NC machine or by a casting or molding, you can rely completely on the CAD model.

Until then, I want my stuff to go out as a drawing, with tolerances. I want it to go out on paper or on PDF so that there is an exact image of what I saw on my computer screen. I am not 100% confident in DXF.

There is an ASME standard on 3D models, but I do not know how many shops and inspectors understand it, and have the tools to support it. I am also curious about how you would mark up your dimensioned 3D models with a highlighter as your check them or inspect the parts.

JHG

 

[ajack1]

With regard to point one why would you want to produce complete manufacturing drawings if all parts can be made and measured to 3D data?

Obviously you need some drawings to specify material, datum points, revision level, tolerance and the like but complete manufacturing drawings?

As someone who also works in the automotive industry for many years now all we get are solid models and the very bare minimum of information, it works fine.

Unless this information is needed for manufacturing or inspection why create it? As far as I can see all you are doing is creating extra work, making the company less profitable and putting peoples jobs at risk, I know I would resist the changes you are trying to make.

I am sure other will see things differently.

 

[ctopher]

Somebody has to check the parts to something. Very rare for a company to have inspectors that understand any CAD software, especially 3D, that can check a model against the real part. If a part file is sent to a vendor, how do they and youe know the part is correct? Does someone open the file in SolidWorks and check each dim in each sketch? Not all dims in the part are created per how you want it machined. This is why dwg are needed. The dwg indicates the rev, dims, tol, title, signatures, notes, etc. Not all of this info is indicated in the part file most of the time. I have always sent PDF's of the dwgs to vendors along with a parasolid, IGES or STL for them to machine from.

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[ewh]

1) The times are definitely changing. Now that more and more manufacturing centers can read CAD data, fewer drawings are actually needed. ASME Y14.41-2003 DIGITAL PRODUCT DEFINITION DATA PRACTICES is the governing standard for those that choose to rely more on the model and less on a drawing. While most CAD packages are not yet fully complient with the standard, they are moving that way. Any information that used to be on a drawing can be in the model file. Granted, it seems most companies are fighting to keep their drawings for reasons mentioned above. However, with the right software, drawings will be unnecessary.
Sorry I can't help your argument on this point.

2) The most effective way to instill this discipline that I am familiar with is to have a knowledgeable, dedicated checker who is given the authority to reject poorly drawn drawings. Checking against applicable standards is his only function, and his word is law. While there will be much grumbling that the changes he requires may be unnecessary, in the long run the drafters (and designers) will thank him, as creating a proper drawing becomes second nature to them.

 

[ctopher]

The problem is, most see CAD as an easy way (and fun way) to create 3D models for machining. A lot of managers, marketing and sales see it as "magic". I am 100% for correct discipline to create files without dwgs, but it will be a slow process. For now, I feel dwgs are necessary.

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[ewh]

You are right, Chris. I was just trying to point out where things are headed. To accomplish that, a lot of training and discipline will be needed, as well as CAD packages which are 100% complient with the new standard. Until then, drawings still have a place in industry.
I don't know mmakm's whole situation, so I wanted to give him a heads up on what's out there. Given that they don't yet follow Y14.5, I agree that they should start with the drawings.

 

[ctopher]

I understand. We are on the same page.
This subject hits a nerve with me.
CAD and 3D modeling is a power tool. I wish it were used to it's full capacity. Unforunately its not in most cases.

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[mmakm]

Thanks for all of your quick responses. Good points in all of them. A few things I need to clarify.

1. All though brought up and trained a certain way, I'm not at all interested imposing my beliefs, just influencing so that the solution makes sense. I'm not at all interested in creating extra work that doesn't provide value to the product.

2. Its kind of comical but sad. We can call out all the dims we want, all day long, but other than a set of calipers, mics and a height gage, we don't have any inspection equipment.

3. We produce in relatively small volumes. Typical EAU is less than 1000 at most and typically in 100 to 300 range.

We had a checker at one point and just due to lack of incoming drawings to check, he was pulled in another direction.

I almost need to look at each part on case by case basis and take into account who the vendor is and thier capabilities. I think being a relatively small company with all business function ocurring in one location, that we have a tendency of taking standard practices forgranted.

 

[ProEDesigner00]

I am working on such ideas for my new company. I fall back on experience from other companies I have worked for that have migrated toward 3-D models for inspection and manufacturing information. Some of these companies allthough say they can go to ANSI Y14.5.1 and prove they are doing things right they still do a crappy job of producing a piece of paper reflecting their intentions.

While being in the opportunity you have to make policy for your company.
1. I would first investigate the successes and failures toward was going 3-D
2. I would look at the cost benifits of not producing paper. This alone can be enough of a saving to keep your company going to the overseas world for cad and design work
3. I would set up a pool of people (former checkers/draftsmen) to become model auditors. These people would be proficient in the use of the cad system to be able to know how to instruct the proper application of the required GD&T and tolerencing requirements. These people would have disciplinary power that if after a set amount of time some one was habbitually not following modeling best practices that these vio;ators would be disciplined after a gracious training/implementation period.
4. I would set up by a small group of software guru's and company product lead people as to what best modeling practices should be. Make a formal company standard of this that has pictorial examples and definitions of what is expected.

I worked for a company that was headed in this direction in most of the multiple facilities they have. It worked very well. It weeded the more unproffesional slugs out of the system as well. I am trying to emulate this type of phylosophy in my new one horse company that will soon be hiring others.

P. S. These points of thought may grow as I ponder what is being spoken about here

 

[ctopher]

Contact someone like Mattel and see how they do it. They do solids from Pro E and SolidWorks and send the files to China for molding. Ask how do they do inspection.

http://www.mattel.com/index.asp?f=true

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[mmakm]

ewh & ctopher,
I agree with you, I feel like I've taken a step back in time. The questions is getting up with the times and how to drive that change. Training is definetly the answer but is not always feasible (for a number of reasons). First, to train I need funds. To get funds I need an argument. To support my argument, I need information on how this will impact the company. I'm sure its the same in alot of places... I've taken, as well as yourselves, the initiative to learn more about the topics that affect my career and industry, but I'm sorry to say, I've seen a large number that haven't or aren't interested. (Just a little ranting & raving, sorry) That becomes the hurdle.

 

[ctopher]

I agree. The people that are not interested...you will have a difficult time trying to change them. I tried, and almost pulled my hair out doing it. Some changed a little, but are unwilling to grab the concept. Good luck.

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[ewh]

I am fortunate to be employed in a small but rapidly growing company. I am the defacto drafting department (with contractors coming and going), and as such, I set the drawing standards for the company. If an engineer asks me to check his drawing so he can send it out to be made, I will do so, but I try follow the standards (I learn more about them all of the time). I often tell the engineers to either redraw it, schedule it so that I can, or get any references to ASME Y14.5 off of it. So they'll send out sketches. Fine by me (and my boss), as long as they accept responsibility for it.
Many of our customers send us models and "reduced dimension" drawings that only specify tolerancing and other critical information. The model is regarded as the master definition. We do the same when outsourcing tooling. It has worked fine, so far.
Whichever path you decide on, good luck!

 

[ajack1]

This subject always seems to get people going, myself included.

To imply that the technology will be in place some time in the near future is wrong, it is there now. For any body panels, facia or complex 3D formed parts the likes of BMW, Audi and Ford simply give a model and a very basic drawing. Everything can and is machined and measured to this, if you do not have the CAD software or the CMM and know how you do not become a first tier supplier, that is how they work now.

That does not mean that this is the best way for everything, if you want a 10mm steel bar saw cut to a length of +/- 5mm then a fag packet sketch and a steel tape is probably all you need, to spend ten of thousands on computer software and a CMM would be a waste and probably see you out of business.

In the same way a shaft with many diameters and shoulders that are highly toleranced is probably best drawn in 2D.

A highly complex 3D form is best modelled and in fact probably cannot be dimensioned or measured using 2D prints.

It all boils down to what you are trying to achieve.

 

[ctopher]

I don't think anyone here implied the technology will be in place some time in the near future. We all know it is here now. The problem is getting more people to understand and acknowledge it. IMO, 2D will be in the past someday, 3D is the way to go now.

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[ewh]

ajack,
If it was one of my posts that you were refering to, I wasn't trying to imply that the technology isn't yet here, just that industry is very slow to accept it. I agree that if you want to be a prime supplier, you stay on the cutting edge.
It IS the practice with some of the aerospace giants, and we have recived packages where the only paper involved was the instructions on how to interpret the model. The only reason that was required was because the CAD software isn't yet fully complient.
I fully agree that there are some parts that just can't be fully defined on paper, short of having hundreds of cross sections. I work with these type parts everyday.
As you said, "It all boils down to what you are trying to achieve".

 

[MintJulep]

The heart of this goes beyond the drawing issue. It is really a process and process control issue.

Someone here on eng-tips once wrote something along the lines of:

A sketch can communicate what your want. A properly dimensioned and toleranced drawing communicates what you will accept.

To do away with a proper drawing is to do away with source or incomming inspection. This is not a bad thing, if you know that all of the parts are good.

Step 1 is that you really have to know what you can accept.

Step 2 is that you really have to know the manufacturing process. Speciffically its limits of variation and repeatability, and that the process is, and will remain in control.

Step 3 is to inspect the hell out of a statistically significant number of "first articles", and calculate the distributions of all the dimensions.

Step 4 is to assess steps 1 through 3, and decide if the vendor has the ability to consistently produce parts that you will accept.

If you can get through steps 1 - 4 then you can have some degree of confidence that most of the parts will be good parts.

Repeat for all vendors.

 

[ewh]

"To do away with a proper drawing is to do away with source or incomming inspection."

Not necessarily. If your operation is set up with the correct software and training, you CAN inspect incoming parts relative to the model ONLY. ASME Y14.41 defines the standards for documenting the features and tolerances of the part sans any drawing.

 

[ctopher]

I agree. But it is difficult to get a lot of companies to use that spec. (as in any other spec also).

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 06
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716