Question
I am curious for your thoughts and opinions regarding managing your work flow when drawing in AutoCAD. This question really applies to those drafting in 2D without the help of a plug in like microvellum and those drafting cabinetry, although this style of work flow can be applied to any drafting. I've seen draftsman for cabinet shops draw two ways in AutoCAD:
1. They draw all floor plans, elevations and sections separately and off to the side from one another, maybe aligning elevations to sections to make it easier to read and line things up.
2. They draw elevations sep. but combine the sections/plans on top of each other and manage model space with layer control either individually or with filters and paper space with viewport freeze.
I think we all can agree that the best way to manage your drawings is to stack sections that are in the same room - cabinetry where you're taking cuts at different heights like upper cabs and lower cabs. Similar to how you would draft plans for a home or commercial building, stacking the floors on top of another. Separating them can lead to errors down the road because you may update one and forget to make changes to another and unless they are stacked it’s not particularly easy to see those conflicts.
So down to the real question, how do you find yourself accomplishing this work flow - is it through layer management or are you using xref's to stack on top but still manage all of those sections/plan parts that are on top of each other? In either case would you mind explaining how you do it in brief detail? I'm always curious to find better and more efficient ways to draft the shop drawings.
Forum Responses
(CAD Forum)
From contributor V:
That is a very good question and there are several possible answers. Let me first say I have worked extensively with both systems and the outcome is largely dependent on several factors. First, how many people are making drawings here and what are their skill levels. A large group of people where a number are less skilled in layer control, viewport layer visibility, etc, the harder it is to pull off stacking of sections/plans.
The other factor is the complexity of what you are drawing. If it’s mostly standard boxes I would say don't bother. But a large number of people out there don't bother with a plan section for simple casework anyway, the thought being that if the guys on the floor cannot figure out how to put a basic box together, inclusion of a plan section for them is the least of your worries. For complex items, with a skilled group of drafters and someone good to pull the system together, it becomes a good idea. You need good layer control for your annotation though, and in my opinion you should not annotate/dimension in paperspace if you are taking this approach as you will lose what gains you make if you do. I also would take the time to create methods that will save time. For instance, take a view port of an upper plan section with the layer visibilities set for that view and make it a block. It is easy to insert it, explode it, and then turn it on in a macro (it will come in turned off).
Then a few set of buttons for basics like dims and labels that set the appropriate layer (like upper detail linear dimension or plan view qleader, etc). Set up this way makes revisions faster, and coupled with dynamic blocks and a good detail library with included dims, labels, etc. already on the correct layers and you have yourself a system.
We recently have branched into cad utilization consulting and this is the sort of project we find to be very rewarding as when complete it is a very efficient and profitable way to create your drawings. The clients are usually astonished at how much faster it is to create and revise very involved drawings for custom work this way.
It also came to me after I wrote the first post that I forgot to include something that is applicable also - how others are managing their face frame and panel breakouts. I prefer to show them off to the side dimensioned in a simple line drawing and keep any details/dimensions for each of those on those breakouts and not on the elevations as it becomes to cluttered if done that way. I still haven't managed to find a better way to draft the breakouts other than on top of the elevations and shifting it off to the side so that I can show them in a sep. viewport. I thought it would be nice also to stack those faceframe breakouts on top of the existing elevations so you could see conflicts when changes are made. I haven't been able to find a solution to how to show them separate when face frames are adjacent to each other and the view port window will cut out some if not all of the lineweight where it is snapped to. Any insight on that also would be interesting to get.
Contributor V - I use palettes to achieve consistent notes, dimensions, etc. on my drawings. The question I always ask myself is what I should make a detail block of. It seems every job I'm doing something different. In the last year I have not drawn the same detail twice. So, is it worth it to create a library of details? I'm interested in not reinventing the wheel, but the wheel that I didn't reinvent still has to fit the car I'm working on.
One thing you may want to consider is annotative dimensions. They are a bit strange to get used to at first, but may be exactly the tool you need for this sort of drawing style while allowing you to not need many dimstyles and dimlayers.
We have one client in particular that is extremely high end custom residential work that is revised endlessly and we switched to doing this. It proved to be much easier and more accurate in the long run. It took a while to get his CAD standards worked out and a fair bit of time to make the macros to automate things, but the time saved made it well worth the time spent.
Set annoscale.
Apply current annoscale to selected dimensions.
Turn off annoscale for selected dimensions.
Create dimension with current annoscale only.
Create dimension with all annoscales applied.
There is nothing magical in these macros and anyone can write them if they are willing to do a little research, but it takes the multistep process of adding and taking away applied annoscales for each object and makes it a click, select, done process. When you can do that, it makes it pretty darn quick compared to the default way of doing it.
