Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.


ANSYS Discovery Live - how does it work? (in general)

ANSYS Discovery Live - how does it work? (in general)

Hi All,

ANSYS just issued new software, please see Link. (I am not reseller, I am not working for ANSYS).
As one may see looking at the marketing video, the new idea in ANSYS is to be able to modify model geometry on run and see stress results plots change in the same time.
I am wondering how it is possible, in other words how they managed to ensure proper meshing and how it applied to nonlinear iterative calculation, in fact why it can be so fast?
After all I do not suspect that they figured out some new algorithm of numerical matrices solving... Perhaps it is marketing only and this "live" instant change-results picture is limited only to some particular examples or geometrical strict configurations.

Eng-Tips Forum Users, please share your ideas and understanding on this new product of ANSYS.


RE: ANSYS Discovery Live - how does it work? (in general)

Marketing talk is a perspective of the reality; it isn't the only perspective, of course. The businessese word for this vision is "democratizing of simulation".

The so-called "reduced order" models are the way to do these analyses. In the coming five years, these models will begin to transform HOW simulation gets done and WHO does WHAT. So, what you saw was not really meant for expert computational analysts; it is meant for the "designers" who want to "see" how the results might change by small changes in the design. But the reason it works is because the meta model (think of it as a simple input-output equation) is created by the expert prior to the designer having done anything.

Are you new to this forum? If so, please read these FAQs:

RE: ANSYS Discovery Live - how does it work? (in general)

IceBreakerSours, thanks for your input. If this is the case and you could refer myself to any White Paper or any other reference for this meta model explanation it will be much appreciated.

RE: ANSYS Discovery Live - how does it work? (in general)

I sat through a webinar yesterday. It was pre-recorded and they didn't take any questions. From what I can gather:
  • ANSYS rewrote some of their solvers in CUDA to exploit GPU compute power
  • Allows users to alter physics/geometry in real time and see effect on results in "near real time"
  • Structural, fluid and thermal examples were shown - in several cases geometry seemned very complicated
  • No discussion of model assumptions/limitations for any of the examples
  • Mesh generation seems to be completely out of users control
  • They noted that this software is not applicable for all types of problems
  • Described as an early prototyping tool rather than traditional solver
  • Aimed at engineers without any simulation experience or knowledge
  • No comment on pricing or business model
  • Free beta/preview available now, commercial release planned for Q1 FY18
Since the webinar was pre-recorded I presume its available on demand on their website.

RE: ANSYS Discovery Live - how does it work? (in general)

Reduced order modeling in computational mechanics (or, mathematics, in general) is a known area of active research in the academia so you can find a ton of materials just by googling. However, to give you a flavor, click here for a talk by a French researcher at a computational mechanics conference on this subject. Basically, what this whole area is trying to do is to reduce the dimensionality of the problems by finding some abstract function space. Think of it like finding eigenvalue decomposition or Fourier transform to take your time dependent problem in to another space, doing your computation there, and returning back to your original Euclidean space.

I only skimmed through the recorded event online; business-talk triggers my ADD. However, I had brief email exchange with our local ANSYS rep and he agreed with my understanding (as discussed in my first comment above).

Are you new to this forum? If so, please read these FAQs:

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close