CAD thoughts: Was: What it was like before CAD

[gfphoto]

I wonder if in some ancient city one day a few of the best architects were sitting around discussing their craft and a fellow shows up with a triangle and ruler claiming all sorts of time-saving advantages...

My son and I have used a relatively inexpensive 3D parametric (whatever that is) CAD package for the past several years and found a few distinct advantages over hand drawing for making telescopes and related parts.

Typical telescopes combine one or more metals and types of glass, sometimes natural materials like wood and rubber, and plastics/composites as well, so being able to immediately find the mass properties is a handy thing: first to know the total weight and center of gravity, and then to quickly see the effects of potential weight saving measures such as removing and/or using or substituting thinner/different materials.

When the individual parts are drawn they can be assembled and the range of motion of appropriate parts specified so that interferences can be seen as the parts are moved. Since you can rotate the model any way less than obvious problems like not having enough room to get your hand in to adjust something are easy to see. Or something that wouldn't necessarily be a physical interference, like a part that blocks a proper view of say, a finder.

I don't think drawing the parts is that much different than what I learned in HS drafting since that starts out in 2D. It's pretty neat to see them extruded and be able to look at them from all directions like you would a physical model.

The package has it's quirks and it can eat up a lot of time. Without my son's help it would have taken a lot more time to get to even my very modest ability. Then again, that drafting class wasn't instant either and probably would have been harder for me if I hadn't had an intro from my dad.

Gary Fuchs

[DickDastardly40]

I find drawing in CAD different to with a pencil, maybe it's my technique or lack thereof.

By hand I try not to use an eraser, in CAD everything drawn overlength then trimmed to size. I often spend ages drawing a part to find it's in the wrong layer or I have the wrong snap settings, which is not something you encounter by hand.

Is it just me (please don't let it be only me)!

Al

[Larry Swearingen]

Well, let's see... before CAD I could see without glasses and I didn't
have Carpel Tunnel in my right hand. {:>)

Friends don't lest friends use snaps .

17 years in CAD and now retired.

Larry

[rantbot]

CAD is great for engineering documentation. No dicking around with vellum, erasers, etc, or trying to find the original drawing when changes have to be made. And in 3-D you can rotate the part to wow the natives - it's cheaper than giving them beads - and you can do the lines and surfaces in colors to impress management. Management always loves colors.

But it's a definite impediment to the design aspects of engineering. The sort of people who can design machinery can already draw - the two skills are strongly linked. But the ability to draw doesn't map well to CAD - the keyboard/mouse interface is all wrong. The better people get at CAD, the worse they get at the more visionary aspects of design.

A few months ago the manager of an engineering department at a small company here in Massachusetts was grousing to me about the sort of help he can get. He showed me a typical CAD drawing turned out by his people. It was a 3-D drawing of a bracket, an elaborate weldment in 6061. It would have been maybe a $600+ part to fabricate. The guy who "designed" it knew which buttons to push to crank out a 3-D drawing, but he didn't know spit about weldments, or 6061, or tolerancing, or structural mechanics. If he did, he'd have sketched out a bent steel bracket costing about $10, and drawn it on a piece of notebook paper in the time it would take the CAD system to boot up.

Progress? Maybe yes, maybe no. I still keep paper around for design work (not graph paper - plain white is the way to go.)

[Michael Moore]

That bracket reminds me of an article I saw in an engineering magazine. They took a somewhat generic part specification around to different shops to get recommendations on how best to make it.

The machine shops said it needed to be machined, the fab shops wanted to weld it up, the foundries said it had to be cast, etc etc.

I like to do drawings, and CAD is a lot more convenient than dragging out the drawing table and finding where my Mutoh and all the different scales are hidden at.

Being able to punch a button and get the mass, area etc numbers can be very convenient.

I see no reason why good drafting practice shouldn't be the first thing taught in a CAD class. Drafting is drafting whether you are using a pencil or a computer.

cheers,
Michael

[Rustybolt]

Some things you can do faster the old way.(Providing you have the equipment) But as far as veratility and making changes go, CAD is better by a mile.

[gellfex]

I design widgets for show business. Before CAD I had a graph paper spiral notebook (still do) and a Casio scientific calculator watch to do trig on. In 1990 I promised a client that I would do a complete set of drawings of the job I did for them. I dreaded doing it on a board so bought a copy of DOS Designcad and learned how to use it. (10 years later that client was still designing puppet mechs with cutout paper dolls stuck together with pins) I still sometimes do a preliminary sketch in the notebook, but get on the computer pretty fast.

I find CAD helps me design by eliminating blind alleys sooner than something conceived on paper. It's the discipline of working at scale. But the program does matter. I found Solidworks too hard to "think in", to develop and throw away ideas, and stuck with Designcad, but I've just started using Sketchup, and find it real easy to think in, especially after adding some plugins.

[tattoomike68]

I found I can draw 8 to 10 times faster with CAD. I took all the CAD classes there were in College. Out on the job I never used it one time.

On the job I would just use graph paper and a pencil, real quick and dirty.

Steel suppliers give you tablets of graph paper free so thats what I use.

[loose nut]

Cad gives you a guaranteed accuracy level you don't have with paper, it will tell you when two or more parts are trying to fit in the same spot, and best of all with 3D solid modeling, you can model a part so complex you will have no idea how to make it.

[oldtiffie]

Don't wurry

I find drawing in CAD different to with a pencil, maybe it's my technique or lack thereof.

By hand I try not to use an eraser, in CAD everything drawn overlength then trimmed to size. I often spend ages drawing a part to find it's in the wrong layer or I have the wrong snap settings, which is not something you encounter by hand.

Is it just me (please don't let it be only me)!

Al

Don't wurry Al.

Its an "age" thing(y).

Having no lead in ya pencil is just part of it!! Having your pencil like a soft eraser is not much help either. I can tell that they play hell with the "Insert" procedure.

I got confused with "Snap" settings as I thought is was to do with the elastic part of things. Seemed to be more "Stretch" than "Snap" there and knowing where to draw the line was essential.

Having to "snap to it" is part of being in the Navy.

"Having/putting it in the wrong layer" is an occupational hazard if you get too pis*ed on a "run ashore". You can (get your?) end up in some strange places and positions.

I tried doing all them sketches in the "Karma Sutra" - I was tracing it by hand on the screen -got terribly distracted though.

Drawing a horizontal line on a ship in a heavy sea-way is a bloody big ask too.

But it must have done some good with the "Volvo" car designers who only designed "square cars" (and encouraged those blasted "Volvo drivers"). They took their graph paper off them and gave them plain paper instead. I suspect that they stuffed it up their AutoCAD "paper-port".

Perhaps its just me.

[New chips]

My Dad was in advanced engineering at GM in th early 60s. He told me that to do a wheal jounce layout would take a weak of drafting then they had to build a test model to try it out in. Now days it would take maybe 2 days to draw and then the computer can test it out and check for problems plus they now have programs that can tell them how making changes affect handling. The other side of the coin is with the old method they guessed on the loads and then overbuilt it. Now you can design right up to speck. A good example my brother had a 68 half ton Chevy pickup with a utility box on it. In the 80s decided to upgrade his pickup and bought a 3/4 ton heavy duty pickup and put the utility box on it and it bottomed out the suspension he had to put air bag suspension on it to.

[gellfex]

My Dad was in advanced engineering at GM in th early 60s. He told me that to do a wheal jounce layout would take a weak of drafting then they had to build a test model to try it out in. Now days it would take maybe 2 days to draw and then the computer can test it out and check for problems plus they now have programs that can tell them how making changes affect handling. The other side of the coin is with the old method they guessed on the loads and then overbuilt it. Now you can design right up to speck. A good example my brother had a 68 half ton Chevy pickup with a utility box on it. In the 80s decided to upgrade his pickup and bought a 3/4 ton heavy duty pickup and put the utility box on it and it bottomed out the suspension he had to put air bag suspension on it to.

That's all great as long as the material science and finite element analysis is correct. I read a book about structures that described how the early 19th century Scots railroad bridge engineers did it by seat of pants, and mocked the French engineers who used math to optimize their designs. Far more of the French ones fell down, because the understanding of metal fatigue was in it's infancy.

The old "overbuild it" had a lot of grateful wartime passengers in stuff like the C-47 and jeep carrying way over design load. I'm only a moderate aviation buff, didn't the C-47 perform at like twice the ceiling and payload it was designed for crossing the Himalayas in the Burma airlift?

[oldtiffie]

"On-site" work

CAD is fine in its element - ie with a computer handy.

I would guess that quite bit of "hand-sketching" is done in many cases before the CAD system is used though.

CAD is not much use if you are on a job where there in no computer or CAD as you will have to "detail" the work by hand on a sketch-pad.

Its bad enough when a job can't be made from a CAD drawing. Its worse if there is not enough or incorrect detail on the sketch to draw it in CAD and a return visit to the job, site, shop - or worse - client - is required.

I would guess that many shops on this forum either don't have a CAD system at all or if there is one that it isn't in the shop.

[brian Rupnow]

Interesting thread!!! I started out in 1965 with a big engineering company in Belleville, Ontario, Canada as a "detail draftsman". I worked through a 5 year apprenticeship, which included 2 years of rotation thru the machine shops, where I worked on a Bridgeport Mill and a lathe for 6 months, on a welders bench for 6 months, a fitters position building gearcases for 6 months. and 6 months with field crews installing machinery. I did all my work on a drafting table for 26 years after that, for various big engineering companies. By 1996 I had worked myself up to a middle managemant position in a company which then collapsed under me, and found that I had to return to my original (and preferred) employment of designing machines and automation for industry. And ----found that no one was interested in my services. It became apparent very quickly that the market perception had become "If you can not design it on a computer, you must not be any good and we don't want to deal with you". So, at the ripe young age of 51, I returned to college, and learned to use computers and to design with Autocad. I worked with Autocad 2D for 3 years, and slowly began to realize that the world had indeed moved on, and now the market perception had become "If you can not design it in 3D with full parametrics and associativity, and then animate it, you must not be very good, and we do not want to deal with you" So--With an investment of many thousands of dollars for computers, software licenses, and training, I entered the world of 3D design. I have been designing with Solidworks 3D now for 8 or 9 years, and I love it!! I will also be the first to agree that knowing how to run a computer does NOT make someone a mechanical designer, a draftsman, nor a design engineer!!! Brian

https://www.rupnowdesign.com