Weigh in: Do All Jobs Require Drawings??

I work in a research support machine shop that functions essentially as a job shop. Management is looking for ways to save wasted billable hours, and the latest is to direct the shop to avoid making drawings unless necessary. In general samples may be provided, and I have been "wasting time" making CAD files before turning to the shop to make the part, or assembly. Please comment, that's what I am here to get, under what circumstances do you not need a drawing of some kind to generate a machined part?

Much beyond a plain bushing and you could be getting something differnt than you expect.


----------

Think Snow Eh!
Ox

I suppose if they hand you an item, tell you it's XYZ alloy, and needs to be "just like this one, +/- .005," then you could skip the drawing... as long as the sample isn't damaged.

Chip

It all depends on your job, but if you aren't the one making it then I will say yes.

Are you in a University setting?

"karnak" did not need a blueprint..........

Since you're in a research environment, perhaps inventing some equipment for the enhancements of telepathic abilities might help with the proper conveyance of requirements....

Anyone with competence in a cad system should be able to generate a model in minutes, depending on complexity. The drawing is usually automated, generates in seconds... I don't see how management will save billable hours. The time that the machinist takes to check, then recheck, then verify the dimensions from the part sample to the new part will easily cover the CAD time. At the end of the process you have a document that shows what you produced.

And also at the end of the process you'll have a drawing/model on file for when they come back in a month and hand you the same part and say "we need another one of these and since you've already made one before we expect it to be done in half the time."

cheers,
Michael

When it's a piece of sculpture and doesn't mate with any other part. Almost every part has some dimensions which are more critical than others. Presumably in your case the person asking for the part may not even know that.

Drawings are contracts and define the part. Solid models are just for managers to look at and "Oooh and AaaH'. I can't see through solid models to check internal features and interferences.

It might be quicker but a some point you will end up with expensive scrap. Use it as a teaching tool for the person requiring the part.

Chris P

bcstractor said:

Drawings are contracts and define the part.

Click to expand...

True enough when considering _one_ aspect of their use.

bcstractor said:

Solid models are just for managers to look at and "Oooh and AaaH'.

Click to expand...

Easy there... Ever hear of "Feature Based Machining"? We use solid models for machining every day.

In our shop, repairs and rework/mods are often done on verbal instructions from the customer. The owner does get them in writing, however, before any work begins.

On edit, after rereading the original post, I realize that's not exactly the same as "generating a machined part".

Just how DO you replicate an old wornout shaft when NONE of the dimensions is original? It IS worn out, remember?

If you don't know the original dimensions, and the mating bushing or whatever, are also worn out. Where do you start?

I could "guesstimate" that, in my home shop, I could recut, if necessary, but I am not working to make money, I could pull the bushing, if there, and mike the OD, and if it is 3/4, assume the ID is 5/8, so the shaft should, also be 5/8.

Bad part is you are talking "generating" machined parts. Keying things in. How do you measure the kink in the shaft, to determine just how MUCH kink it has to have?

Do you HAVE the original manual for the machines to read FROM the plans you already have, to key in your CNC data?

The Bosses are handicapping you. I take it you are a keyboarder, NOT an Injuneer, else you would tell them to go pound salt. " I know what the fuck I am doing! Kiss my ass! You don't like it, FIRE me!"

As an afterthought, it is PART of Management, but really the IE who is the prick you have to challenge. All them bastards do is cut, cut, cut. You gave them an inch the last time study, they want a foot, this time.

In retrospect, tho' I made good money, 30 or so years ago, I actually ran machines that were charged to the job at 100 per hour or more.

I was paid (say) 12 bucks an hour, and my incentive was 1.50 per hour. SO, I did what we called "200%", 16 hours cutting in 8 hour shift. So my hourly for that week was 13.50, and the Corp billed twice the hours in that period of time, that week.

Wage, I got 12 out of 100, 88 for the Co., double the production, I get 1.50, Co. gets 98.50.

Seems unfair, I suppose. Way it was. Steel mill, 20% bonus was an actual 20%. Steel mill is still there, Westinghouse is just a Japanese memory .

Cheers,

George

I worked in an instrumentshop located in the middle of lots of labs for 7 years.
Since changing jobs 3 years ago I`m one of the people having stuff made in the current shop. (again a research environment)

Have to agree with what`s mentioned here : it depends......
In my opinion THE attitude to show in a research environment is flexibility.
Close second is knowing when and how to educate your customers a little.

So I would do the simple stuff without drawing and make at least a sketch for a bit more complicated jobs.
This sketch would indeed save the day when a year later your guy walks in and says : remember that gizmo ? ..... another one please cause I had to send the first one to Berlin.

The hard part is ofcourse deciding when a `bit more complicated` starts.
As also mentioned : a skilled draftsman does a quick drawing in a short time.
So make sure you are or become one.
(completely ignoring acually designing stuff ofcourse , which usually takes more time)

R & d

The better engineers I`v worked for always started out with a sketch and dated it in the event that any patent issues arose. Once the concept was proven out and a working model was produced, the prototype is sent out to the customer for validation. Meanwhile the sketches are sent to the draftsman and put on paper.

So the customer comes back and says we need ten each for testing.

Ok, so the request is sent to the model shop with the newly made cad drawings, and guess what? there`s a bunch of mistakes on the drawings.
So now`s the time to prove the drawing, keep in mind that I still have the sketchs.

The first ten products are made to print. If there are any problems, changes (mark up`s) are made to the drawings. Parts are sent off.

Customer reports back and says, product was vibrated at 10 G`s in three plains, and it fell apart!

Oh Shit! Back to the drawing board.

There are many reasons to document all along the way.

Keep in mind that only 10% of the product usually goes to market.

...MC

paceplane said:

I work in a research support machine shop that functions essentially as a job shop. Management is looking for ways to save wasted billable hours, and the latest is to direct the shop to avoid making drawings unless necessary. In general samples may be provided, and I have been "wasting time" making CAD files before turning to the shop to make the part, or assembly. Please comment, that's what I am here to get, under what circumstances do you not need a drawing of some kind to generate a machined part?

Click to expand...

I used to work in military research as a machinist.

We used to have 2 ways of approaching jobs, first was designers/drafters drawing up a part to be machined, this approach worked fine so long as the science guys who drew up the request never changed what they wanted in the 4 weeks between the request and the job being issued to the shopfloor.
The other way we went about it was that the machinists would be assigned to a science group (I was S17's pet metal basher), the science guys would tell me what they wanted, do a few sketches on the back of a cigarette packet and go machine the stuff, If it worked then we'd pass it over to the drafting office so they could draw it up for mass production by sub-contractors.
We preferred the 2nd method since we could talk and see what the science guys wanted, and easily modify the parts if there was a problem
Plus we got a trip out to fit the equipment to whatever ship it was going on.

In conclusion I'd say that not every part needs a full up works drawing to start with, especially in the prototype game, but does if you're going to mass produce it.

Boris

I often make things for myself without a drawing or sketch of any type, sometimes multi-part stuff of some complexity. OTOH, I often need to make my own CAD drawing just because it's easier to have everything nicely dimensioned and without mistakes.

That said, most people aren't very good at communicating what they want, if they even know, and I'd never make much of anything without a drawing, even if it's a quick sketch on a napkin. The only exception might be the walk-in request, "can you take 5 thou off this surface?"

IMO, a good solid model person can work very fast. I'm not him. Solid models can waste huge amounts of time, or they can be very useful. Depends on the situation. Most of the time 2D drawings suit me fine if they're properly drawn, i.e., used as a form of communication to the machinist. Seems obvious, but a lot of people seem to think drawings should be a mystery to be figured out.

Making a solid model for one broken lamp detail (chamfered outside, two different bores) - it took about 5 minutes including printer warming.
Would took much longer for machinist to make the drawing by hand.

And if the part is very easy and it does not need a drawing - some notes are mandatory anyway: ie. "pipe, D=10mm d=6mm l=250mm 10pcs", because even the time it takes to walk from engineering room to workshop can make you forget some things.

So the time savings which your management is looking for may well be "negative".

Thank you all. Every response is appreciated. Any more posted will also be read. Thanks again.

Drawings, long and short of it...

Long:
I was newly hired in a university research machine shop. A gentleman who quickly became one of my favorite PhD's. First of all he calls and makes an appointment, he arrives with a scale model of helicopter hub assembly, a beautifully crafted thing.
He tells me he wants to redesign some components, we sketch on a chalk board, I make some suggestions that he likes, some that just will not work and he explains why. We wrap up with me saying that I'll make some some dimensioned drawings and review them with him before starting the work.
"Oh no he says, we will make the drawings including the details we have discussed and present them to you for your review - "The machine shop must be shown the proper respect, a machinists time is very valuable, we must not squander it on being too lazy to create a proper document."

I grin broadly and shake his hand - you and I will get along just fine!

Short:
Here's a thought, if time and paper is so valuable then why don't these bean counters stop writing down their budget, just remember it and verbally pass along the information!

I believe a lot of time and money can be saved by only making basic prints or none at all.

A lot depends on what the machinist is capable of. Some machinists can't do anything without a 50 page print that specifies every last detail. There's nothing wrong with that, but those types of guys are better suited for production or say, building complex aerospace parts.

In your research support shop you need a different type of machinist that is comfortable interpreting the intent of the designer. There may not be time to make proper prints and the designer may not even have the skills to in the first place. Or they may even slow the whole job down by making prints with overly tight tolerances.

For example, I don't need a print to tell me the surface finish, straightness, squareness, diameter tolerance, or whatever else for a shaft that needs to fit a bearing. I know the intent of the engineer is for the shaft to fit a bearing and support X part. Maybe I cut the shaft and it has some taper but the diameter is correct where the bearing is going to fit. I'm done instead of starting over because it didn't meet some unnecessary print specification. The engineer gets his parts, does his testing, and is on to the next revised design. All that matters is that their design intent was met.

In conclusion, it all depends on the machinists, the engineers and what sort of relationship/work environment they have. Loosey goosey napkin prints and verbal instructions can be an awesome way to work. You can get a shit load of work done in a short period of time. On the other hand, you can also get burned big time when the design intent wasn't met and you have to start over.

In a modern research support shop that has CNC and is working off solid models, there should be a lot less need for detailed drawings in my opinion. There should be just enough information for the machinist to understand the design intent and then take it from there.

Where I work the engineers greatly appreciate not having to always make detailed drawings. It saves time on their end and mine.