Optical / CAD software
Helpful person
optics and CAD? The name is OptisWorks and can be found at:
http://www.optis-world.com/OPTISWORKS/optisworks_cooperative.htm?soft_id=3
Its advantage is that it is integrated into the CAD which is useful
for setting up parametric models. (I suspect that it works better
than their web site.)
Boxman
> Is anyone familiar with an add in for Solid Works which integrates
> optics and CAD? The name is OptisWorks and can be found at:
>
>
> Its advantage is that it is integrated into the CAD which is useful
> for setting up parametric models. (I suspect that it works better
> than their web site.)
I haven't had any direct experience with optis works. I've had some
time with their demo for the SPEOS software integrated into Catia V5.
SPEOS is the illumination modeling package that Optis sells. It is
nicely integrated and worked well. If I recall, there were some
technical differences such as using meshes to approximate the surface
from the CAD to do the ray tracing with, so your accuracy depended on
how fine a mesh you used, and it handled ray splitting differently
than I was used to. But ease of use seemed to be good and being
integrated directly into the CAD package made it easy to work with the
mechanical engineers if you used Catia. Ultimately, the price was the
biggest problem as they were very expensive at the time (this was a
few years ago) if I recall it cost nearly $70,000 for the package that
we would have needed.
Helpful person
Thanks very much for the info. They do not have any prices on their
web site. Your quote seems extremely high.
My questions are because they are giving us a demo next month and I am
trying to get some advance information.
JB
"Helpful person" <rrl...@yahoo.com> wrote in message
news:10ead9d3-f953-4dad...@r37g2000prr.googlegroups.com...
There was an online Webcast demoing this package yesterday from Optis. Very
powerful tool indeed. We use Solidworks & Photopia here but we are looking
seriously at Optisworks despite its ?25k price tag (on top of the Solidworks
seat too).
JB
Helpful person
>
> - Show quoted text -
Interesting. Apart from the quality of the software I have concerns
regarding the stability of the company. We are getting a demo of
their software and nobody told us about this on line webcast. In
addition they seem incapable of putting up a decent useable web site.
(I spent months on mine.) $25K seems a very high price. I'm
surprised that they don't follow the Zemax model of inexpensive
software and pricey (fairly) maintenance/service agreements.
Thanks for the information.
Boxman
>
> - Show quoted text -
JB,
If you are doing illumination design with solidworks as your CAD
package, I would also look at Light Tools from ORA. They just
released their solid works link module that gives you a linkage
between your solidworks model and your optical model so that updates
in either software are reflected in the other.
http://www.opticalres.com/news/lt61_slm.html
Light Tools is a good piece of software and ORA is very responsive to
it's users. (BTW I don't work for ORA).
Out of curiosity, if you don't mind, why are you switching away from
Photopia?
Jamie
> Is anyone familiar with an add in for Solid Works which integrates
> optics and CAD? The name is OptisWorks and can be found at:
>
>
> Its advantage is that it is integrated into the CAD which is useful
> for setting up parametric models. (I suspect that it works better
> than their web site.)
There really isn't a good add-in for standard CAD packages that
provides the functionality of 3D CAD constructs and optical
interaction modeling. You need to determine in which discipline the
technical risk lies and choose your design/analysis software best for
that. I find it unlikely that most mechanical design engineers will
understand the difference between such specialized definitions as
radiometry and photometry.
If the application requires metrics best define by optical
definitions, then the application should probably give the best
methods for defining, analyzing and optimizing those requirements.
Generally, optical design and analysis programs such as Code V and
FRED provide the best tools for defining and opimizing parametric
models for optical requirements and constraints. These applications
provide native support for modeling optical field distributions and
field/surface interactions. Code V gives better optimization
capabilities for sequential and mostly sequential optical models. FRED
provides analysis and optimization for complete mechanical structures
and rays tracing that is based entirely on non-sequential search for
surface geometry intersections and interactions. Both allow the
problem to be defined, optimized and analyzed using optical
definitions and requirements.
The optical system's mechanical model should be easily translated to
an appropriate format for mechanical CAD applications to place and use
for subsequent parametric solid modeling for packaging the optical
design. I use SolidWorks for this phase of the design effort. The
optical elements are replaced by Solidworks parts that are
parametrically defined from the optical model and then features are
created to mount, center and align the optical components. After a
preliminary model of the opto-mechanics is completed, the entire
system is returned to FRED for comprehensive modeling for potential
stray light issues (EMI for you RF types) and simulation for design
assurance.
Unfortunately, there really is not one application/add-in combination
that can design, package, model and assure optical system designs. But
be very happy that the gap between the mechanical design and optical
design, modeling and analysis, software capabilities has become very
narrow. The best example of this is FRED from Photon Engineering. FRED
provides for objects based on optical design definitions while also
providing the ability to natively define and create mechanical
surfaces as well. FRED is a relatively new comer to the market and
hopefully will inspire other companies to lower the bar between
mechanical engineering and optical design.
James A Carter III
www.opticalconsulting.com
Boxman
I respect your opinion here, but I disagree with what you are
saying.
Optis along with Light Tools and TracePro are at a much higher level
of integration between CAD and Optical than FRED. Optis is fully
integrated into the CAD system and Light Tools and TracePro offer
packages that directly connect the CAD to the optical analysis such
that the parameters defined within the CAD model (i.e. sketch
variables and relations) are conserved and can be changed from either
package. Also optical features such as light sources and surface
properties are carried in the CAD model for use by the optical
modeler.
Packages like FRED have no such linkage.
Any non-sequential analysis package has the ability to create both
optical geometry as well as mechanical geometry either by direct
definition or through translation (or usually both). It's the direct
linkage to CAD that Optis, ORA, and Lambda offer with their packages
that offers a higher level of integration than other packages like
FRED. I'm not saying that Optis is necessarily better at developing
optical systems because it is integrated with CAD, I haven't used it
except for a demo, but to the point regarding integration, it is well
ahead of any other package except for maybe the recent release of
Light Tools solidworks connection. Programs like FRED, ASAP, and
ZEMAX are well behind the curve in terms of CAD integration.
No matter the level of integration, it still requires an optical
engineer to perform or at least oversee the analysis. Having the
analysis integrated directly into the CAD like Optis does, or having a
good connection like ORA offers speeds up product development
tremendously and reduces by many times the likelihood of errors
occurring when translating a design from the optical analysis package
to the mechanical package.
JB
>Out of curiosity, if you don't mind, why are you switching away from
>Photopia?
We have no intention of switching away from Photopia. I'm still an AutoCAd
user. All of the Sloidworks seats are in our luminaire design dept. I run
R&D and don't really get much involved with the mech designs, only the
optics/reflectors/light sources etc.
Photopia works fine for my needs.
JB
Boxman
Thanks JB. I have a similar situation here, except I use ASAP with
Rhino in R&D and our design department uses Pro Engineer. There are
copies of photopia available, but I haven't used it for any design
work. That's why I was wondering.
jaca...@earthlink.net
Just curious as to what you do, optical design or mechanical design? I
guess I can appreciate that you might like to have parametric
mechanical designs that you tweak that include optics that can be
modeled in the mechanical package but I have know of very very few
people that understand and use tools like SolidWorks and also do
optical design and modeling.
Further, most of the detail in large mechanical assembly models are
not needed for the optical modeling and simulation. This spurious
detail only slows the non-sequential ray-surface search and slows work
flow (similar to SPICE modeling used by electrical engineers, they
rarely model the whole circuit board, just the important networks
unless they are doing RF designs and parasitics are important and then
they're not using SPICE). For example, it is highly unlikely that a
chamfer or fillet on the _outside_ of a lens housing will affect
performance and some mechanical designers include fasteners and other
components that have little or no relevance to the optical
performance. I guess that you can suppress these in the CAD portion of
the "integrated" package as I do when I export the mechanical model
for import to FRED.
Note that I work mostly with laser illumination, scanning and sensor
applications. If one works with illumination and lighting systems, the
mechanical CAD package may be he best place to "edit" the optics.
However FRED has the ability to model just about any arbitrary surface
profile, it's easy to do this in FRED as well but the model parameters
will be the ones that an optical engineer is more familiar with.
Again it comes down to your training and your orientation or
discipline. Mine is mostly optical and my optics drive the mechanical
design for packaging with precision. So generally, the optics are
defined first and then used in the mechanical parametric design for
packaging. I usually do all the metal-glass interfaces and layout
folds, etc. A much less expensive mechanical engineer then details the
design and packages it within a product design. FRED is used at
various points for design assurance and stray light assessments, etc.
Rarely does the optical design change or need to go back to the
mechanical model in any way so that direction of integration is not
very important.
I saw posts that mention Optis is something like $25k. That's enough
for 2 FRED licenses and 2 SolidWorks seats as well, with some left
over to play with. I would like to meet the user that finds that the
high level of integration of Optis, TracePro or LightTools (which is
leased not even sold) is worth that difference in cost. Correct me if
I am wrong here, but if you collaborate with another engineer, doesn't
that person need a license too, in order to keep all of the optical
properties matched to the mechanical features as well as the
parametrics for the mechanical model intact or can anyone with
SolidWorks play with the file and not ruin the optical aspects of the
model?
BTW FRED does something that none of your packages can. Optical field
synthesis with coherence so interference effects are properly modeled
if a sufficiently experienced user creates and manipulates the model.
I assume that these others can do ordinary thing like polarization
effects and volume scattering too...
If I had a employer that was willing to dole out $25k for another
tool, I might be tempted to try one but for now I am extremely pleased
with FRED and how it plays with SolidWorks. But I have a degree in
optics and have been doing this for over 25 years.
Just my $0.02
Boxman
Hi James,
I have a degree in optical engineering and did a lot of work in
college with holographic interferometry, but when I got out into the
job market the defense industry was tanking and the only ones hiring
were the automotive companies to do illumination engineering (Oh how
times have changed), so I have spent the majority of my career working
with illumination optics. I also spent a couple years working at
Breault Research in the engineering services department and as an
instructor for the illumination courses. Hence my preference for
ASAP.
You probably know this, but ASAP was the first optical analysis
software to approach non-sequential raytracing needs outside of the
classical optics arena and automotive lighting design was the area
where it got a lot of application. As you pointed out, different
areas of applications need different levels of tools. In automotive
lighting, we ended up needing to be able to model entire assemblies
including the surrounding "non-optical" CAD geometry because stray
light is a huge concern especially with headlamp systems due to the
regulatory requirements applied to these systems by the federal
government, and we needed to be able to accurately predict
illumination system performance to avoid needing to do costly
prototype tools. A non-sequential monte carlo ray tracer was the only
way to do this. ASAP was the first (at least that I could tell) to
integrate CAD based geometry capabilities into their software by
having several bezier and polynomial based geometry definitions with
the ability to bound and trim these entities with edge base objects in
the core modeler, and they were one of the first to build a CAD
translator to get CAD directly into their software.
If you have ever looked at the optical systems in automotive lighting,
you know they aren't standard optical shapes found in classical
optical design programs, and they rely heavily on CAD based
descriptions. As CAD evolved, the manufacturing capabilities evolved
with it, and the ability to do more and more complex shapes for better
illumination performance has evolved with it. So for me, CAD and
illumination design have long been intertwined and I have had to
become proficient in both. In order to get a good illumination design
from end to end, I need to understand the description in the optical
software, the description in the CAD software, the description inside
the manufacturing software, and ultimately the plastic processors
setup and execution. (95% of the illumination optics I have done are
made from plastic). I use Rhino as my CAD program, because it is a
hyrbrid modeler (surface based modeler that creates closed solids)
that exposes the inner working of the Nurbs that it is using clearly,
and I can control the information flow across the process, like when
it goes to a CNC machine to be cut, I can verify whether the CNC
program has done any approximations or if it is retaining the Nurbs
description. If it approximated, I can simulate that approximation
with Rhino and ASAP and see if it has any detrimental effect. The CAD
description in ASAP is identical to what I do in Rhino so I am
confident that my CAD model being simulated is what I will eventually
be tooling. Also since ASAP's CAD description is script based, it's
extremely easy to optimize a system using ASAP's buit in optimization
routines (downhill simplex and simulated annealing) or custom
algorithms (I use a genetic algorithm most often for illumination
design).
Tooling and manufacturing for classical lens optics for example,
follows a completely different path, and doesn't require a CAD based
description.
Having said all that, if you view the CAD system as an application
environment that can be expanded as the CAD companies have begun to
do, then you can see how you could build optical descriptions into
your CAD program and perform both the optical and mechanical design in
one environment. The CAD vendors have built their platforms using
object oriented design, and thus have opened up their core to
application development through SDK's. it's now possible to
instantiate your own features with their own functions that can be
managed by the CAD system as though they were it's own. So you can
use all the display, rendering, file handling, etc. capabilities of
the CAD system without having to do all that programming and just
program your object's special functionality. You could, for example
then define a spherical lens object that was equation based, and then
write a custom routine to draw it for the CAD to display, but any time
an optical calculation was run on it, it would use the equations of
your object, and not the CAD polynomial representation. This is
transparent to the user, so it looks like a seamless application.
That is where Optis is ahead of the curve from other companies. They
have built their software to reside inside the CAD program, which
makes it a seamless application (for some cases) of both mechanical
and optical modeling. There is probably more drive to have CAD and
Illumination optics combined because the two are already more
intertwined. These are the users that are driving the integration
development that Optis has undertaken and Light Tools has just begun
to pursue. I know of at least 3 companies that use optis because of
it's integration, and my compatriots that use Light Tools are thrilled
with the solidworks link and claim to not ever want to go back to an
unlinked package. I'm assuming that as people see the benefit of
having the tools integrated (whatever they might be for them
personally), the classical optical design packages will eventually end
up moving to CAD applications as well. I think, however, that optical
engineering is such a specialized field that it will take a long time
for that change to occur because there is a limited market to chase
and making funds available to rewrite programs into a CAD environment
isn't going to be easy to do.
Plus I'm a die-hard ASAP user and don't care for the implementations
I've seen so far of the other programs because they all end up being
constraining compared to what I can do in ASAP. So they probably
won't get my dollars either. I appreciate your enthusiasm for FRED
and have considered it at times too since I know the pedigree of the
developers there (they were at Breault before starting Photon), but
I'm too invested in ASAP and the price difference isn't significant
anymore.
As a side note, to be fair ASAP models coherent systems with a similar
implementation of Gaussian Beam Decomposition as FRED. I understand
the appeal of FRED on the price and ease of use side though.
Perhaps more than you wanted, but that would be my overly long $0.02
worth!
jaca...@earthlink.net
III. Just FYI...
I appreciate that you have a totally different discipline and
perspective for optics than I. What you may not appreciate is that I
pay out of my pocket for the tools I use and that cost is very
important to my personal bottom line. Further, I know the the folks at
Photon Engineering quite well. I worked and been friends with Rich
Pfisterer long before he joined Breault. I have also used ASAP but
that was at least 8 years ago. I found it unnecessarily obtuse and
extremely poorly documented and avoided nonsequential tools like it
for that reason.
Until FRED was released. After purchasing a license, I took the 3 day
FRED course where I was joined by 12 other people. All of these were
moving to FRED from ASAP which goes beyond coincidental. I have used
Rhino as well but only the earliest versions. It appeared then that it
was intended for people creating 3D objects for virtual reality
frameworks and was not intended for precision mechanical design, much
like trueSpace or AutoDesk's Maya. That may have changed in the years
that have past.
You mention "stray light is a huge concern especially with headlamp
systems due to the regulatory requirements applied to these systems by
the federal government.." These are probably less onerous than the
performance requirements of federal military standards regarding
optics based avionics but that is not important (stray laser radiation
can actually cause catastrophic failures in the optical subsystems).
What I might mention is that a certain large military contractor where
I was once employed moved completely away from ASAP and into FRED
mainly for productivity reasons. A friend of mine, that is still there
with well more than a decade of ASAP experience, was actually happy to
leave it behind.
However, as you state, your work flow is entirely different than mine.
So I understand that optis is probably your best choice. Perhaps I
will look farther and attempt to get a demonstration. This will
probably be difficult as it's hard for me to convince their sales and
marketing folks that I have pockets deep enough (or a business case
strong enough) to actually buy it.
I wish you good luck and fortune and hope that these evolving tools
make your tasks easier to handle. One thing to remember is a quote by
a well known optical researcher and professor where he predicts that
the world will need only one optical designer due to the advancement
made with computers and software to optimize a set of centered
spherical surfaces. The moral is that as the tools advance, so do the
problems we need to solve.
Lucky for us!
Jamie