Drafting Pratices?

Just wondering how others do there layouts? In the old days (R14) I used to
draw everything in model space, one big layout then used layermanager to
turn the correct layers on/off. I also had the titleblock in model space as
well with the page info on different layer (i.e. G-ANNO-PG01, G-ANNO-PG02).
This worked well and kept everything in one file. I have seen people use
different files for different sheets and I have also seen many layouts in
model space with titleblock in paperspace and use layout tabs for each
sheet.

What are some of the advantages to these methods, are people do other things
as well? I am evaluating different approaches and I am looking at making
things easier, if there is a better way.

This is for arch, but other disciplines may be doing it a better way.
I also use a modified AIA layering standard.

Any advantage to the new sheet sets? How is it used?

Thanks

--
_________________________________
Timothy Spangler

"You cannot escape the responsibility of tomorrow by evading it today"
Abraham Lincoln
_________________________________

AutoCAD 2002
WinXP
Compaq Evo W6000
Intel Xeon / 1G RAM

 

Right now we create a base map in a single drawing. Then we Xref that to a
drawing that will contain our proposed objects. We then have all the
proposed linework and objects in model space. Then we crate one layout for
each sheet. We are a Civil / Survey firm and we find this works best for us.

If you are considering upgrading to 2005 at any time. I would suggest you
get a trial version and evaluate the new Sheet Manager before making any
revisions to your procedures.

Allen

 

I can only speek about what works for us. I'm sure there are others out
there more qualified than myself about Arch but we do our drawings in
model space. All of our stuff is small so there is only two formats that
we use and that is the A and B size formats. When a new drawing needs
generated the user selects the format and that opens a blank drawing,
sets all styles, layers, line weights, standards and loads the
titleblock in paper space. When the user needs to modify the titleblock
then they must go to paper space and edit the attributes. Notes are also
placed on paper space. As long as the drawing is drawn 1=1 in model
space to scale things up or down is easy in paper space.

Hope this helps
Mark

 

Beat to a pulp here:

http://discussion.autodesk.com/thread.jspa?messageID=3529824&#3529824

 

One can get really complicated here which can cause a lot of extra work. One of
the world's largest oil companies which I worked for a few years back used one
drawing (model space) for each discipline, i.e., structural, piping, electrical,
architectural and if there was to be any combination of one or more disciplines,
they were xrefed. Title blocks were drawn 1:1for a myriad of scaling factors,
i.e., 1/8"=1'-0", 1/4"=1'-0", etc, and inserted at the scale of the drawing,
i.e., if 1/8"=1'-0", the title block was scaled in at 96 from 0,0.
Title blocks had attributes and the company had a huge number of blocks with (or
without) attributes which would be inserted based upon the scale factor.
This gave the company one drawing for each elevation and caused the least amount
of work and/or files.

 

This gave the company one drawing for each elevation and caused the least amount of work and/or files. <<

That's an opinion not shared by all.

 

All I said was this was ONE method - and they are still the world's largest oil company.

 

There is no single 'way' to do CAD drawings, there is, however, one single
'way' to approach each project...

The most important thing to consider is the project - how you do your
drawings depends on how the project needs to handle your drawings.
It also matters if the work is 2D or 3D, the end-product is paper prints or
a CAD file.... let's put it this way: if your end-purpose is paper drawings
for installation or record, then you can be as sloppy as you like with your
CAD (just get it done and plotted). If it has to integrate in some way with
other people's CAD systems, then you have to be consistent. (no strange
fonts, linetypes, dimstyles, layering ideas etc)... A CAD project meeting is
essential right at the start - so everyone can sing from the same songsheet!
*********
For example, let's say there's a central resource for drawings - for
co-ordination and approval - perhaps an architect or main contractor. Then
each subcontracted trade (for example, small power, heating, ventilation,
steelwork, glazing etc) would have to work to some grand-scheme, some
consistent method for xrefs and layers.

There is little point in everyone returning architectural information to
source, so everyopne would xref this stuff and detach it before returning to
the central point for co-ordination and approval.

Similarly, there is little point in everyone's title blocks messing up the
central resource, so it makes sense if this sort of thing goes on a
tab/paperspace.

This means that when all subcontractors work consistently for the PROJECT,
their (relatively tiny) drawing files are received at the central point for
co-ordination and approval and xref'd into the architectural stuff. Clashes
are spotted, layers controlled, progress made!

This becomes all-the-more important on large projects with many levels of
sub-contractors!
*********
On small jobs with no subcontractors, or for quotations etc. it may be
judged easier to take the file sent, create the layers you need, open a tab,
pop in a viewport controlled by layers, slap in a titleblock, and put all
your text and dimensions on the paperspace tab, but draw directly onto the
architect's model. Then return the file to source.

This has the advantage of being contractually consistent revision-wise.
*********
Huge companies - who have been using CAD badly for years - tend to be slow
to change tried and tested routines born many years ago! Some have in-house
CAD packages that are unavailable elsewhere and so forth, and this is
especially true of oil-related work. It must be remembered that such huge
companies work to different rules - their CAD guys are often
freelance/agency personnel as well ... and this means they will find the
slowest and most complicated way to do everything to justify their
timesheets!
*********

 

That is sort of how I used to do it at my previous employer. Drew all in
model space I had written a Tilteblock program that inserted the titleblock
at the correct scale in model space and layers were controlled by stated
i.e. FloorPlan, Pluming, HCAV, Electrical etc.. This seemed to work because
it was only drawn once in one place. The only draw back was if you had to
add a changed and you needed a new layer for it, (say revision A) a new
layer would be created for each revision well this screwed up the layer
stated you had to go back through each state and resave it.

--
_________________________________
Timothy Spangler

"You cannot escape the responsibility of tomorrow by evading it today"
Abraham Lincoln
_________________________________

AutoCAD 2002
WinXP
Compaq Evo W6000
Intel Xeon / 1G RAM

 

All I said was this was ONE method - <<

The part with which I took exception was the part where you said "caused the least amount of work and/or files"

It would increase the number of files required over using multiple layouts in single files thereby requiring MORE files. And it is my opinion that working with the same data spread over several files is considerably more work than woking with that data in one file.


Which means....??

 

For THEIR (repeat, for THEIR) use, this was the best, causing the least amount of work and/or files. Tolerance is the ability to see another viewpoint, i.e., to each their own.

 

We do structural drawings on 24X36 and 30X42 sheets. We have a separate
model space drawing for each plan level and for each sheet of sections and
details. We then xref the model space drawing into a paperspace file which
contains the titleblock. If we need to overlay another disciplines plan
into our drawing, we do that in the model space drawing.

The reason we chose not have one file with all the drawings in it is because
then you limit the number of people able to work on the files. We use a
numbering system for the xrefs so we know what sheet they are xrefed to and
the scale factor of the xref. This has work fine for us.

Different strokes for different folks.

 

I can see where someone may prefer that method over another, but to say it creates fewer files is inaccurate.

Tolerance for 2+2=9 is not tolerance, it's just silly.

 

In this part of the world (New Zealand) normal practise is for the title
block to be an xref, usually in paperspace (where it ought to be), but not
always.
Does anybody else do this?
I am currently with a Mechanical company, but have previously worked in AEC
and found the same there.
We have over 60,000 drawings in our database and only one title block for
each paper size. The actual title text is a block with attributes.
Aside from the obvious disk space savings, another advantage is that we can
update our address details on all drawings in a few minutes.
Our GA's work much like AEC (we make Timber Drying Kilns) in that we have
separate sheets for Foundations, Structure, Cladding, Pipework, Electrical,
and so on.
Each sheet is a separate drawing, for legacy reasons. And we do not use any
xrefs (much to my dismay) except for the title block. Which means when the
client does something like reverse the grid numbers, or change the
foundation, we have to edit every drawing.

Now layers........hmmmmmmmmmmmm........
If a line is yellow and continuous, it goes on a layer called "y-con"
If a line is white and hidden, it goes on "w-hid"
I think you can work the rest out......
I think its called a Pen Based Layer System.


Having made a fuss about the "Drawing board mentality" of our CAD procedures
as a whole, and backed up by some negative customer comment, I have been
given the joyous task of proposing a new system, using xrefs, blocks, and
intelligent layer names.
Actually I pleased to be doing it, because its a rare opportunity to get
things right. All I need now is the time to put proper thought and planning
into it.
Obviously I'm keen to see what others have to say about their practises as
well

 

Also, less to lose if a drawing becomes corrupt if separate drawings are
used.

 

Hi, Kev;

Yes, that's the way to deal with TB. It has been proved times and times again to
be the most efficient way.
Here is how I name layers. Once the model is finished the first drawing created
is General
Arrangement with item balloons pointing at individual parts. After that I open a
Model file and
create as many layers as Item balloons in the GA. The names of layers correspond
to the numbers in
item balloons. Say, I have items 1, 2, 3... etc. Then in Model file I will
create layers 1, 2,
3...and put corresponding parts of the model on their layers. I end up with a
lot of layers
(depends on the size of the project) but it isn't a drama. I can always suppress
unneeded layers
with layer's filter and leave visible the ones I work with.
Using this approach to layer names allows me to access quickly any part and
view it individually.
If a client calls and ask some questions about, say, Item 25 (that's the item
number he sees on
the printed drawing) than all I have to do is to open the file and turn off all
layers but 25. And I don't have to think in a hurry, what layer did I put this
item on?
Please keep in mind that the system has been developed for 3D drafting. However
I think it can be adapted for 2D drafting as well.

As for the Standard convention of layer names for Mechanical Drafting I haven't
seen one written
yet. Those I have came across with were company standards and applied to an
individual company
only. Hence, they are not standards in broad sense of this word.

Regards,

Igor.

 

Yes, that's the way to deal with TB. It has been proved times and times again to be the most efficient way. <<

Oh? And the data to back up such "proof" is...?

If you require a block in the drawing for the editable attributes of the titleblock anyway, there's is little advantage for the TB to be an XREF.

We may have 6 different contracts for the same client, different locations running at any given time. Title blocks as XREFs provide us with no production advantage.

 

I am sure you know that Xrefed TB can (and should) have editable attributes in it. If it isn't the case than working with TB is a two stage operation, which isn't bad
either. First the border gets Xrefed into the drawing, than TB information is Inserted and edited as needed. By the way, I always wander why most of the times the TB
with attributes has lines in it, if you see what I mean. Why TB lines cannot be a part of the border and get xrefed with it? My attributes for the TB are just
information, no lines.

As for the advantages of the Xrefed borders over inserted ones.
1. They take less memory in the drawing, that's a proven fact.
2. Borders usually do contain information relevant to your company,- address, phone, E-mail, logo e.t.c. If the logo should change due to some reason, than to update a
million files with Borders inserted into the drawing is a full time job. With borders (and title blocks) xrefed it is a matter of a few minutes. Example - Telstra
(Australian telecommunication company) changed its logo one time and it should be reflected on all of their drawing. The change was implemented without any problem.

 

Thanks for the intellegent reply Igor.
As a matter of fact our DMS attaches the tb as an xref and inserts a text
only block with attributes when we press the 'new' button. We can enter, and
edit title block data, including parts list, without opening the drawing.
I've not worked at a lot of places but its the best system I've come
accross.

I take it you're using somthing flasher than basic ACAD for your modeling.
When I do the occasional 3D job, draw the parts, xerf them into the sub
assemblies, then xref the S.A.'s into the G.A. I shorten each xref name to
the 3 digit sheet number, which is also the part number, so it has a similar
effect to your method. My newish machine isn't spec'd for 3d work so I dont
have the processing power to put everything in one drawing.
I also use numbered colours (11 and over) for my 3d layers, and named
colours (10 and under) for everything else, so I can sort by colour when I'm
setting up my layer states.

 

1. They take less memory in the drawing, that's a proven fact.<<

If you require a second block for attributes you've gained very little file size advantage. When the bulk of our models weigh in at over 7 Megs, 10K for a TB isn't even noticable.

million files <<

And that change does not impact legacy files. In fact, in many cases can not impact them legally. The company for which I currently work, has changes names a couple of times over the last 12 years. Contracts originating under the old name were required to maintain the old name for liability issues.

TB should be set in stone at contract, not doing so leaves one open to constatnt TB manipulation that is not cost productive.


Poor planning should not an excuse for using TB's as XREF's. But I understand that we can not control our client's whims (unless you've set it in contract) A simple logo change can be implemented and Automated nearly as easily if the TB is a block in the drawing. A company hired us for a major contract and after contract was sold to a different organization who desired a new TB. We used a lisp function (in place for other uses) in batch mode to redefine the block in all drawings, took less than an hour with modt of that going to building the new logo.

I still see little advantage to XREF'd TBs.