Drawing in layer 0

[Hardeight]

OK, we are in the process of restructuring our drafting standards. One guy, who is relatively new to CAD, is trying to draw everything in layer 0. I tried to explain this was a bad idea and he asks why. And to my surprise, all I can come up with is, "that is what I have always been told."

I could go into technical details about all the weird things layer 0 does with blocks and xrefs but is there any other real reason why we shouldn't draw anything in layer 0?

Or is it just and old CAD wives tale these days?

[David Bethel]

There are certain times drawing geometry on layer 0 makes perfect sense. But not if you need any type of layer control or manipulations. Esp if you use paper space. It all depends on your discipline.

 

Are you an architect or engineer type? -David

[alanjt]

If something is on the 0 layer and that file is XRefed. Everything on the 0 layer will take on the layer properties that the XRef is placed on.

[David Bethel]

If something is on the 0 layer and that file is XRefed. Everything on the 0 layer will take on the layer properties that the XRef is placed on.

Same with a standard block.

 

So lets say your are making a title block. Why not make everything on layer 0, control plotting line widths, linetype, etc with colors. I can't think of many times you would need freeze / thaw title block layers in paperspace tabs. -David

[Patrick Hughes]

There are exceptions to every rule.

 

I always create any block geometry on layer 0 but when I insert the block it goes onto a specified layer other than 0 and will inherit those properties.

 

One reson for not drawing on layer 0: how are you able to isolate the viewing of certain geometry?

[Patrick Hughes]

Oh, and a good alternative to "that is what I have always been told" or "We've always done it that way" is "Let me think about that"

[alanjt]

Oh, and a good alternative to "that is what I have always been told" or "We've always done it that way" is "Let me think about that"

 

 

Or: "you know, the last several people we've fired, asked the very same question"

[alanjt]

Same with a standard block.

 

So lets say your are making a title block. Why not make everything on layer 0, control plotting line widths, linetype, etc with colors. I can't think of many times you would need freeze / thaw title block layers in paperspace tabs. -David

 

Right, but with a standard block, you want that to happen.

[Patrick Hughes]

LOL very nice!

[Hardeight]

Or: "you know, the last several people we've fired, asked the very same question"

 

That is a pretty good one, I will have to stash that one in the old comeback bank.

 

We are an engineering based office. So sometimes it seems OK to have a lot of things on one layer. We have so much information on some of our drawings that you will literally have to come up with hundreds of layers to keep them all separate.

 

I am all for simplicity but it just seems like the obvious reason of no selection or visibility controls would be reason enough to make more layers. But our problem is coming in where do you draw the line. We have so many special cases that it is almost impossible to dictate how to do everything in every given situation. If you leave our drafters up to their own devices you will easily end up with with thousands of useless layers.

 

My cohort wishes the extreme opposite. I am trying to find a happy medium. His idea is easier to document as a policy but mine would be more useful.

 

Herein lies my dilemma...

[ReMark]

Put the policy into place as an Engineering/Drafting standard and enforce it with the CHECKSTANDARDS command. To do this for a group of drawings use the Batch Standards Checker.

 

There are plenty of people looking for jobs. If your co-worker doesn't want to play by the new rules book show him the door and find a more willing candidate.

[alanjt]

Stick around, I'm full of snide remarks.

 

That is a pretty good one, I will have to stash that one in the old comeback bank.

 

We are an engineering based office. So sometimes it seems OK to have a lot of things on one layer. We have so much information on some of our drawings that you will literally have to come up with hundreds of layers to keep them all separate.

 

I am all for simplicity but it just seems like the obvious reason of no selection or visibility controls would be reason enough to make more layers. But our problem is coming in where do you draw the line. We have so many special cases that it is almost impossible to dictate how to do everything in every given situation. If you leave our drafters up to their own devices you will easily end up with with thousands of useless layers.

 

My cohort wishes the extreme opposite. I am trying to find a happy medium. His idea is easier to document as a policy but mine would be more useful.

 

Herein lies my dilemma...

[Crazy J]

If something is on the 0 layer and that file is XRefed. Everything on the 0 layer will take on the layer properties that the XRef is placed on.

 

I think this is an important comment. I am trying to organize my assembly drawing by using x-refs. I am finding that if done properly, it greatly reduces the file size and allows me to organize my layers more neatly.

 

 

 

 

 

 

 

Here's my general approach thus far (and I still have a lot more to learn).

1. Come up with the names of the various layers (one for each component or component type in my case).
   
2. Draw the new component in that layer. When I am satisfied with it, export it to a new file using WBLOCK command. Give it a descriptive name (I am putting 'block - seal -cover' for example) and I have been storing them all in the folder with my assembly drawing for ease. This will probably have to change as I start working on new assemblies/products and want to re-use these files.
   
3. When I am making the block, I be sure to choose a logical base point, and be sure that in the main drawing there is a point that I can osnap to later when I re-insert the block as an x-ref. I also choose "delete from drawing" under the objects section of the dialogue box.
   
4. I then go and re-open the block I just made, making sure it is how I want it. I chose 'select all' and move everything to Layer 0. I then make Layer 0 current and delete all other layers. If I don't do this, I will get a bunch of junk layers imported into my assembly drawing and it will cause the file size to grow (rather unpredictably as well). If it was a tough component to draw and I might have other junk in there, then a PURGE command might be appropriate as well. Also, I typically choose the top view so that I can have a consistent approach for putting the X-REFs into the main drawing. Save this file.
   
5. Finally, I go back to my main drawing and make sure that whichever layer I want this component on is set as the current layer. I use the XREF manager to bring the component back in. I do a quick check of my layer manager to see that I don't have a bunch of new layers created that were unintended. If there are no new layers, then I know things worked as intended. I hit a save again and check my new file size and compare it to the .bak file size. Again, if that hasn't jumped all over the place, I know I am good there. I then copy, mirror, or array as needed. Save, and I should be good to go to move onto the next component.

I've written this to share my method with the forum, but also to hear critique on my method. I'm still pretty new to all this, but I see the value in learning XREFs, so please let me know if this approach holds any water.

[Pablo Ferral]

Standardising 'Supporting Geometry' layers is pretty easy, ie. Centerlines, Hidden lines Dimensions, Annotations e.t.c. Standardising the content can be much harder. It's probably better to come up with a system and allow users to create their own layers ,than try to get to rigid about it.

 

Toolpalettes are your friend, if you make it easier to use the standard than to do it there own way you can win them over!

[alanjt]

Standardising 'Supporting Geometry' layers is pretty easy, ie. Centerlines, Hidden lines Dimensions, Annotations e.t.c. Standardising the content can be much harder. It's probably better to come up with a system and allow users to create their own layers ,then try to get to rigid about it.

 

Toolpalettes are your friend, if you make it easier to use the standard than to do it there own way you can win them over!

 

Ditto!

I like the way you think.

[rkent]

Why use named layers?

AutoCAD gives us multiple ways to manipulate the data that goes into a drawing file via layers:

We can freeze layers that we might not want to show or not show in certain viewports.

We can assign line weights to layers to better show our intentions.

We can assign colors to layers and linetypes to better show our intentions.

The line weights, color and linetypes can be changed on a single layer but then you have lost the flexibility to change them to something else easily on the fly.

With mulitple layers I can lock all layers but one and easily manipulate one layer that represents a system, object, etc.

When plotting, the layers can be plotted in different colors, lineweights, linetypes per viewport (latest releases).

I find layers invaluable when working in 3D. Freezing layers that clutter the drawing area make it easier to work. And plotting 3D is easier as layers that you don't want to "get in the way" are easily controlled.

We can have layers showing but not plot.

Having multiple colors makes it easier to keep track of different parts, systems, etc.

[Patrick Hughes]

 

 

When I am making the block, I be sure to choose a logical base point, and be sure that in the main drawing there is a point that I can osnap to later when I re-insert the block as an x-ref. I also choose "delete from drawing" under the objects section of the dialogue box.

 

I then go and re-open the block I just made, making sure it is how I want it. I chose 'select all' and move everything to Layer 0. I then make Layer 0 current and delete all other layers.

 

I take a similar approach as you with an exception or two.

 

When I create a block rather than selecting a base point I always use

0,0,0 as the base point. Then when I open the block drawing I just made to move objects to layer 0 I also move the geometry to a logical position in relation to 0,0,0. This way I never have a problem with the BASE ir INSBASE sysvars.

[Tankman]

Oh, and a good alternative to "that is what I have always been told" or "We've always done it that way" is "Let me think about that"

 

"and, I'll get back to you later." :wink:

 

Never going to happen.

 

I think a lot depends on what your drawing, and the program you're using.

[Crazy J]

When I create a block rather than selecting a base point I always use

0,0,0 as the base point. Then when I open the block drawing I just made to move objects to layer 0 I also move the geometry to a logical position in relation to 0,0,0. This way I never have a problem with the BASE ir INSBASE sysvars.

 

I think that makes sense to me, but I need to read up on those sysvars...

 

One reason I ended up with this method was because when I set up my original drawing, I probably don't know where my 0,0,0 really is. I mean, I can find it, but I was learning AutoCAD 3D at home at nights as I was learning the product during the days and trying to start my solid model. The guy who hired me knew I only had ever done 2D, but since I'd been trained in 3D SolidEdge, he thought they were all the same and it should be no problem. Lots of late-night hours later, I'm getting a handle on things. But when I started this whole assembly, I didn't know of half the components that were in it or the many others that the other guys keep adding to it

 

If I keep at this and get kept on for another project, I will take another couple steps in creating a standard work approach that is a bit more logical. I suppose I could go in now and see about choosing a global 0,0,0 point and relating them all back, but for now I'm running short on time after messing around with XREFs.

 

In automotive (my former life), everything was "in-car-position", much the way you explain it.

[Crazy J]

I take a similar approach as you with an exception or two.

 

When I create a block rather than selecting a base point I always use

0,0,0 as the base point. Then when I open the block drawing I just made to move objects to layer 0 I also move the geometry to a logical position in relation to 0,0,0. This way I never have a problem with the BASE ir INSBASE sysvars.

 

I was at work today and your comment about 0,0,0 base point clicked in my mind. I was making a new block to bring back in as an X-REF and it was like "duh".

 

So what I realized is that the trouble I have taken to carefully pick a base point and then make sure I load it in at the right spot was a bit of a waste of time for single-use components, particularly as I am developing the design of the component right in the solid model file of the entire assembly. For things like screws or spacers that get used in different spots or so facing up and so down, for example, I would still say that the center of the underside of the face is the spot that makes the most sense to use as a base point, as that is the surface that will mate with whatever you load it into...

 

Anyway, I tried the 0,0,0 with the block I was doing and it worked well and saved me some time. I am nearing the end of this solid model, so I am not going to go back and change things. But should I start more projects, I will better utilize this approach.