[Search for users]
[Overall Top Noters]
[List of all Conferences]
[Download this site]
Title: | EUCLID |
Notice: | CONFERENCE MOVING TO NODE KEPNUT 12/3/89 |
Moderator: | KEPNUT::LAMOUREUX |
|
Created: | Wed Oct 12 1988 |
Last Modified: | Fri Jan 20 1995 |
Last Successful Update: | Fri Jun 06 1997 |
Number of topics: | 43 |
Total number of notes: | 147 |
32.0. "SLA how to do it from Euclid" by CUBICB::BARKER () Thu Nov 16 1989 12:51
---------- Building Euclid parts on a 3-D Systems SLA machine----------------
-----------------------------------------------------------------------------
by Dr. Chez
& Dr. Al
First, let's start with some important safety tips.
1. The available building area of the SLA machine in the Mill is
a 9" cube. Your part must be small enough to reside within
this cube.
2. The part must reside in positive space. That's positive X,
Y, and Z. It's reccomended that you place the part about
.020 off the X and Y axis, and about .25 inch off the XY
plane, hereafter referred to as "the floor".
Both of these are key safety notes, as they keep you from building
the part right onto the sides of the tank, avoiding nasty cleanup
work.
3. When you do finally run the Euclid Slicing algorithm on your
part, Do it in a session with ENGLISH UNITS and a 9" WORKSPACE!
Create the part in metric if you wish, but slice it in english.
Using english units simplifies things for the people running the
SLA machine, as this is what they're used to. Using a 9" workspace
simplifies the downloading of the resulting STL file.
Now you can't be TOO dangerous. Let's move on.
-----------------------------------------------------------------------------
Take a good look at your part. Are you satisfied with the faceting
of it's curved surfaces? I hope so, cause in the words of the im-
mortal Flip Wilson, What you see.. is what you get!
If your curved surfaces are too coarse, a couple of methods for
upping the number of facets follows...
The first method is to file your part, and do a computed retrieval
in a new session with a minimum workspace and a polygonalizational
coefficient of 2. If your part is *VERY* rough, this computed
retrieval method might not up your faceting enough. Make
your best judgement on this, and use the second method for maximum
faceting if need be.
Computed Retrieval Method
1. file your part, and log out of your present session.
2. get into a new session
In the primary table, set your workspace to the BARE minimum
for your part.
In the secondary table, set that poly coefficient up to 2.
3. Do a computed retrieval of your part.
commands: mgmt/exit, retrieve group, toggle compute to YES!
4. Now, the file you see "should" have twice the # of facets.
5. unprotect, and re-file your part.
(1)
Tree Edit Method, for the maximum # of facets (brown belt users only)
1. Get into TREE EDITING.
2. "display" the contour that you prismed to make your solid.
3. Get out of TREE EDITING. (previous view)
4. up the # of segments within your original countour lines
commands: line, compound line, polygonalize.
Pick your line, and input the # of segments you wish.
The number of segments = number of facets in resulting solid.
5. Once you have done this to all the curves in you contour,
make a "closed contour" of the resulting highly segmented
lines.
6. "prism" this closed contour to the same value as in your
original solid.
7. Get back into TREE EDITING.
8. "substitute" your new prism in, where the old one was.
9. "execute" at the top level of your tree.
10. get out of TREE EDITING.
11. The curved surfaces of your solid should now be generously
faceted.
12. unprotect, and re-file your part.
-------------------------------------------------------------------------
--- Wow! Now you're cookin.-----
All the curved surfaces on your part are smooth, you've got your
part atleast .020" off the X and Y axis, and a quarter inch off the floor.
You're looking good, you're feeling good.
On to the support structure. This is the separate part that
"cradles" your part off the floor. It is a solid piece, like
your part, and extends from the floor, embedding itself about
.020 into your part. One of the handy methods of trimming this
support is to translate your part up .020 in the Z, cut the
support, and then drop your part back down. Another important
note here is that your support piece must cradle all of your
part. If you have any horizontal extentions or bosses off your
part, there must be a rib to support these. If your part proceeds
vertically for awhile and then runs horizontally again, your
support must extend up through your part and support the upper
horizontal area. The SLA builds your part by solidifying
horizontal sections with the laser. If these sections are
unsupported, they will droop over, or perhaps even break away.
The machine is depending upon a horizontal surface to solidify
the next layer onto. This is the function that your support
structure performs.
The Part file, and the Support file are two separate entities.
Do not assemble them into one part. They co-habitate some
amount of space, but are separate entities. By being separate,
The SLA operator may adjust their densities independently of
each other, making the support structure into a egg crate type
of solid, and your part an entirely solid entity.
--------------------------------------------------------------------------
That, in short, is most of what you'll need to know to get your
part ready for building. Manipulating the sliced STL file is
a whole other thing however. The file has to be converted from
VMS to ULTRIX, and some other manipulations have to be performed.
See part 2, authored by Dr. Al Lamoureux, for file manipulations.
Good Luck!
(2)
----------------------------------------------------------------------
---------------------Dr. Al's File Manipulations----------------------
---------- Or: How to get your file to the SLICE computer-------------
----------------------------------------------------------------------
Once you have followed the above procedures to create the part
in EUCLID the next step is to "SLICE" the object in preparation for
part creation on the Stereo-Lithograpy Apparatus. At the present time
the "SLA" slice interface is accessible under the user applications and
has not been linked directly into EUCLID.
Once everthing has been created ie: part and support structure if
necessary, and positioned in the positive quadrant, exit from your EUCLID
session. Next log into EUCMGR account to access the SLA user application.
At the DCL prompt type in the following: EUCIS SLA/APPLI=SLA. This will
bring you into EUCLID with all the appropriate defaults for the session.
You may want to change the project,subproject,user settings, all other
portions of this initilization file must remain as is.
Retrieve the object and support structure into your current session.
Next select the USER APPLICATION from the applications menu. Select "SLA".
Toggle the output file to "YES" and output fromat to "ASCII" from the SLA
submenu then select the part and "GO". This will create a file in the default
directory of (partname).stl , next do the same to the support structure.
You will now have two VMS files with the .STL extension.
The next operation will convert these VMS files to the proper format
rquired by the SLA controller. Exit the Euclid session and at the DCL prompt
type the following command:
"CONVERT/FDL=VMS_TO_ULTRIX.FDL (INPUT FILE NAME) (OUTPUT FILE NAME)" IE:
CONVERT/FDL=VMS_TO_ULTRIX.FDL SUPPORT.STL SUPPORT.STL
When this has been completed for both the part and the support the next
step will be to copy these properly formated files to the SLA system.This
wil be done by the following DCL command.
COPY/LOG (FILENAME).STL SLICE::(filename).stl IE:
COPY/LOG BLADE.STL SLICE::blade.stl
Note the use of lowercase characters in the second instance of the filename.
This is required for the Unix SLA system. Once both the part and support
structure STL files have been copied to Node SLICE, the SLA operator can
load your files and create your part.
----------------------------------------------------------------------
Another Important Safety Note:
Make sure to call our SLA operator extraordinaire,
Charles Patterson (dtn 223-8748)
or write at FRSBEE::PATTERSON
to let him know you intend to send him this file.
Sometimes he needs to clear some old files out of the way before
the slice computer has enough storage space to accomidate your
file, particularly if it is a large one.
Special thanks to:
DEC PEOPLE
Al Cassista, Laura Nickerson, Mark Buckman, Russ Stanton, Charles Patterson,
Al Lamoureux, Billy Carlson, Charles Barker, and Carol Boudreau
MATRA PEOPLE
Pete Hadja, Gena Jackson, and Tim Illingworth
And all the other good folks who helped to turn this formerly irritating
and highly unsavory problem, into a predictable process.
(3)
T.R | Title | User | Personal Name | Date | Lines
|
---|