Simulate Early, Simulate Often... In Rhino
For me the best thing about Scan and Solve is that Im in on
Version 1 and I get to watch it grow.
I guess I like being a trail blazer.
Bob, as a proud parent, I am not sure how to respond. On one hand, I do not want to brag about my children, but on the other hand I certainly do not mind others praising them. And some well-intended criticism would be Ok too. :-)
More seriously, you are an experienced user of ANSYS and other systems (?) What do you get with Scan&Solve that you do not have with ANSYS? (Other than 95% reduction in cost.) Care to post any interesting examples? Would love to see what you are doing with Scan&Solve. Thanks!
The main reason I purchased this software is that I can own it myself as
apposed to my employer. Its really affordable. Scan and Solve is software
for the people. You are freeing engineers from the oppression of their employers.
I guess that makes your company a symbol of democracy.
When I was in engineering school I felt that technology would reduce the number of jobs
for engineeers, but in fact the opposite appears to be true. Technological evolution provides
us with more interesting work with the amount of time being dedicated to mundane
engineering tasks like drafting being reduced we have more time for analysis, creative
design and problem solving.
I like ANSYS alot. I find the software provides very good results and is widely used.
The company spend lots of money on research.
Quite often it takes a small company that is free from buracracy to come up
with unique product like Scan and Solve.
It is not exactly what the answer I expected. I was hoping to see some examples that are problematic in ANSYS because of meshing. But Ok then, at least we have a new motto for Scan&Solve:
... freeing engineers from the oppression of their employers :-)
I actually get into alot of trouble with things that can't be meshed in ANSYS.
It a problem I run into all the time.
ANSYS iterates alot and attempts at finding a solution to meshing and if it cant find a find a solution
it will flag the area it is having trouble with and then provides an error message. I can look for some parts I
had trouble with and provide a picture of the troubled area. It would be extremely interesting to
see how Scan and Solve deals with the same part.
To solve the meshing problem I usually edit the geometry in SolidWorks then bring the new geometry back into ANSYS for re-analysis. This processs takes a considerable amount of time, sometimes days. I m charging a fair amount for my time to my clients and when the time is eaten up by what I consider software problems. It can become an issue with my final bill to my client.
I find that trying to get a part to mesh is very frustrating, when it happens. I would rather spend my time
doing analysis and providing good service to my clients.
Thanks Bob. Very interesting. Would love to see some examples ... and see what Scan&Solve would do with them.
... so a better motto could be: " freeing employers from the oppression of the software" ... or better yet, "freeing software and users from the oppression of meshing". About two years ago, in an interview with ANSYS magazine, I said exactly that: Scan&Solve was intended to liberate users from the tyranny of meshing. For some reason, the interview did not get published ...
Anthony, thanks very much for you thoughtful and insightful comments. No offense taken. I do not think that anybody was suggesting the compare S&S analysis capabilities with ANSYS which costs two orders of magnitudes more. And of course, we are adding capabilities, as you noted. However, I do not agree with your assessment of meshing as a limiting factor -- in ANSYS or any other mesh-centric system. (You do not either, or we would not be having this discussion.) But your other comments are particularly interesting. There are two separate issues: technology and economics.
Technologically, S&S is an interoperability solution. It can work with any geometric representation and we have other versions of S&S in-house. But to build a standalone version means that you have to import a variety of formats. This introduces possibilities or errors and greatly increases the cost of software. In fact, Rhino is a pretty good and inexpensive importing solution(!) if you want to go this way. On the other hand, one of the unique things about Scan&Solve technology is that it can work directly with native geometric representations in any system, without conversion -- something no other FEA solution can claim today.
Economics presents much bigger challenges. The idea of low cost software works only if there is sufficiently large number of users who will invest in the software. So from this perspective, you are absolutely correct that a standalone solution serving a variety of systems/users is much more attractive. Larger sales mean more customers, lowering the prices, which increases the number of customers, and so on ... unfortunately this chain reaction works in reverse just as well. So which way is it going to go? We are still figuring it out, but it does depend quite a bit on you -- the customers.
And finally, thank you for the pointer to the other systems. We are always on lookout for opportunities and for competition.
Indeed, meshing is not an issue ... as long as it "works". But often it does not. Meshing cannot "work" because, by definition, it has to deal with many unsolved problems of model repair, adapt to noisy boundaries, defeature, and so on. When I say "unsolved", I mean not fully automatable. This is a fundamental difference. As a user, you may *think* that you do not care about it, but this is what makes Scan&Solve what it is.
We did look at EnMesh and Z88 ... but really? Let's just say that we are not terribly concerned ... and I think that you grossly underestimate the cost of licensing components from other companies. This is one of the reasons why software costs so much.
Making Scan and Solve into a stand alone program isn't
that far fetched. Wouldn't you purchase the use of an off the
shell modeling kernel and then build the software on top of
it. McNeal and Associates might not license/release the use of the Rhino
kernel for that type of usage but there are off the shelf NURBS modeling kernels available.
I see Scan and Solve growing and I'm excited to see where it goes from here.
Thanks for looking after the team Vadim.