How to lower drafting/drawing errors 6

[jkdrummer]

We produce mechanical drawings for ground support tooling in aerospace. The number of drawing changes due to either a drawing engineering error or drafting errors has to be reduced as per our leaders.

We are a large company and have standards, checkers, designers and engineers doing it all together. The errors range from wrong heat-treatment specs, parts list errors, to missing dimensions. Our drawings do not require GD&T so it's not rocket science. Most of these errors are found by the fabricators and then we issue a drawing revision. And we get tracked on the quantity of errors and the number of drawings being scheduled and revised. The checkers are naturally the ones to blame, but they work hard and that's all they do all day.

Has anyone else had this task to reduce these error numbers with success?

 

[KENAT]

Yes, but our fix was in part to institute checkers which you apparently already have.

Do the checkers have the level of experience required? thread1103-193286 is relevant.

Do the checkers changes get fed back to the folks making the drawings, so not only does the drafter make changes on the current drawing but hopefully learn for the next one?

You could try and institute a checklist or at least more formal procedure/process for checking. I'm not a huge fan because any detailed process is either going to be very long or leave things out however you could look at this one

http://www.cad-portal.com/articles/art_details.php?ID=17.

However, maybe some kind of training from the best checker of the sort of things to look for and his process. For instance I used to start with the title block, then check notes, then part out lines to see all required dims are there, then mating hole patterns, then other internal features, while looking at tolerances think of the manufacturability... - but I wasn't a very good or very quick checker.

Are the changes checkers ask for being made - do back checks get done to verify this.

You say you have standards, do you use the typical 'red for change' 'yellow for keep' type annotation? (Green for designer comments, Brown for engineer comments, Blue for checker comments & non mandatory changes if you want to go crazy)

Do the checkers have time to do a proper job (this could go for others too)?

Do the checkers get the information they need like mating part details, description of function...

Do the actual drafters/designers/engineers have the required experience & qualification - would they benefit from some more education of some sort?

Do you track the errors in detail? If so perhaps every few weeks you could get together to see what are the most frequent types of errors being made and bring more attention to them or something..

Do you have as standard set of notes that you can customize for each drawing, or does the drafter have to come up with the notes each time?

Our drawings do not require GD&T so it's not rocket science.

- I'll not touch that with a barge pole.;-)

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[CheckerHater]

Missing dimensions and missing parts in BOMs are inexcusable, and checkers are to blame. Nevertheless some percentage of errors is always trickling town to the shop. This is why the manufacturing companies have prototypes and trial runs to debug the process before the final release.

I understand your company does tooling and every product is unique and possibly only build once.
In this case some fine-tuning on the shop floor is inevitable. Your management has to distinguish between “errors” and “follow-up” necessary to document shop-floor changes. ”The number of drawings being scheduled and revised” has nothing to do with quality of your work.

Could it be that you are working for “tail is wagging the dog” company, and manufacturing is just manipulating the design?

You mention “wrong heat-treatment specs”. According to standards you don’t specify the process on your drawings, so you only use “hardness”.
You also mention not using GD&T. All together makes me think you don’t have real standards but rather some bunch of old photocopies that your checkers hold sacred and do not share with the rest of the group.

I may be completely wrong but I’ve seen many places like the one I described. So solutions to your problems could spread from getting “real” standards, “real” checkers who understand them, “real” spine to stand up to manipulative shop to, possibly, “real” management more familiar with the nature of the industry you are in.

Implementing all of the above is another story

 

[ewh]

I agree with KENAT on this one (surprise!). Good checkers can go a long way in improving drawing quality. It is also important that the same person who made the mistake is the one to correct it. The repitition of correcting your own mistakes allows you to strive to recieve a back check with less red marked all over it. This include dimensioning schemes as well as missing dimensions, etc.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[Belanger]

Our drawings do not require GD&T.

How do you control the size, form, orientation, and location of each feature? It's quite cumbersome to do all that without GD&T.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[ewh]

Geometric Dimensioning and Tolerancing covers more than tolerancing.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[drawoh]

jkdrummrt,

You have the processes in place to reduce errors.

How competent are your designers and checkers? If they were given more time, would the increased cost be made up for by reduced screw-ups in the shop?

To me, this is a design checking issue. Do the errors caught by the design checkers cost more than it costs to pay the design checkers? If you were to tie up and gag your designers, drafters and checkers, and lock them in a closet somewhere, they would stop making mistakes.

--
JHG

 

[TheTick]

The checkers are naturally the ones to blame

That's like blaming the word processor because your story sucks.

If you are over-reliant on your checkers, you are under-reliant on yourself to produce good work. Reduce the number of errors for the checkers to catch, and the error rate will go down. Then, if the checkers are not inundated with a flood of dumb mistakes, they are more likely to catch the mistakes that matter.

I also sense an over-reliance on processes and an under-reliance on competence.

 

[MintJulep]

Having bodies associated with the boxes on the organization chart is not the same as having people with the right skills to do the job. Are your people appropriately qualified for their jobs?

Perhaps everyone is overworked and rushed. Even skilled people make errors if they are not allowed sufficient time to do their job properly.

 

[dgallup]

Real CAD systems properly implemented can go a long way toward fixing some of those problems. If you have predefined materials with predefined heat treatments it should be difficult to get incorrect heat treatments specified on the drawings. If you BOM comes from the model it should be difficult to get incorrect BOM on the drawing. Note the electronic drafting tables like AutoMAD are not real CAD systems.

A firing squad should be able to clean up the rest of the errors. (just kidding)

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[ewh]

CAD systems go a long way in simplifying the process, but they don't stop people from forgeting or improperly applying dimensions or applying ludicrous dimensioning schemes, at least not yet.
As I've noted before elsewhere, someone should not be allowed to use a CAD system for drafting until they understand how to create a properly dimensioned drawing.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[MintJulep]

CAD has made the situation worse, not better.

All any unskilled operator need do is pull the trigger on the "Dimension Shotgun" tool and BLAM! There are dimensions everywhere.

 

[ctopher]

Quality drawings have declined over time. The lack of drafting/drawing training is the cause, and sometimes laziness.

Chris
SolidWorks 11
ctopher's home
SolidWorks Legion

 

[CheckerHater]

Don’t blame CAD.
Hiring bums from under the bridges in an attempt to save money is what caused decline in quality.

 

[KENAT]

Yeah a CAD system properly configured and used can help but people round here were producing terrible drawings using CAD for years before we tried to clean up their act.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[ewh]

It is a wonderful tool, but not the only important one in the toolbox.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[jkdrummer]

CATIA: Making bad drawings faster!

 

[looslib]

Don't know what CATIA has, but PTC has ModelCheck and Siemens has CheckMate as a tool that will check certain aspects of a drawing and model. In our case, ModelCheck runs vevry time we save a file and checks for certain parameter and file settings. The user is given a chance to review the messages and make corrections. There are a few settings that unless they pass the check, the part will not save.

You can not blame the checker, depending on their job description, for the errors. The checker can only be blamed for missing a mistake that the designer piy into the drawing. The 'dependent on CAD to do it right' management are the same ones who hire high school kids and turn them loose with a $20K seat of CAd and expect good designs. Engineering is still required to understand the requirements, apply the specifications and create manufacturable design drawings. Training for using CAD is usually treated as an expense when in reality it is an investment in both the employee and for the company to get effecient use from their CAD purchase.

Every race team in the Indy500 uses the same chassis, yet only a handful have a real chance of winning. The team preparation at the shop and in the pits is vital to transform a basic car into a winner. CAD is the same way. Each company needs to customize the system to work for them, automate tasks and be an effective tool in the design process.


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[ewh]

Model Check and Checkmate have been a boon for designers and checkers. They're great for maintaining correct layers, properly filled in title blocks, modeling issues, etc, but AFAIK they still don't address dimensioning schemes, missing dimensions, or tolerance stackups.
Most of us here know that it is difficult to self-check, and unless we have the luxury of putting the drawing aside for a week or so before revisiting it (rare these days), we will miss some obvious problems that we would have caught if looking at it with "fresh" eyes. Peer checking depends on a high checking skill set amoung all peers. This is one area where the checker improves efficiency and is most valualble.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[truckandbus]

We supposedly use a four eye principle before the design is sent for final release. The general complaint with this is each person's workload doesn't really take into account time spent on other people's work. We also have issues with people on a lower payscale finding errors in the work of people paid more. We have a mutual agreement with purchasing and our vendors for electrical harnesses and panels that there will be errors in quantities for small parts like seals and screws; superceded part numbers for relays and fuses, etc and that we let the vendor collect the errors until the first sample is made and then we update the print to address them on one release.