Empathy List Archives

MG

Mohammad Gharaibeh

Mon, Jun 9, 2025 11:39 AM

Dear experts,

I am trying to use some AI stuff in my simulations. I have an experimental
stress-strain data extracted from tensile tests at various temperatures and
strain rates for a solder alloy. I have obtained a relatively good fit
quality for ANAND model parameters (R-sqrd = 80.8%), and have used MATLAB
to train an Artificial Neural Network (ANN) to find the relationships
between stresses and strains, strain rates and temperatures. The ANN fit
has R-sqrd of 93.1%.

What I am after is that I want to conduct FEA simulations using APDL
(Version 2012) to compare the response of the solder joint when exposed to
thermal cycling using both material models, e.g., ANAND and ANN.

Is there a direct way of having ANSYS dive into the saved ANN MATLAB file,
and pull out the stress-strain relationships?

I have spoken to ChatGPT and he gave me a suggestion to use USERMAT
subroutine but that needs some decent programming. Another suggestion, is
to use the ANN to generate stress-strain data into tabular format for
different temperatures and strain rates, and then define such relationships
(discrete data points) in APDL using TBDATA and TB, RATE and provided some
APDL lines for it, and it looks decent but haven’t tried it yet.

Does this sound doable? I mean, I know that it’s possible to define
stress-strains for different temperatures but I am not sure if that’s
possible for different strain rates. Is it?

I would appreciate any feedback and suggestions.

Best,
Mohammad

—-
Guest Researcher
University of Freiburg
Germany

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

Dear experts, I am trying to use some AI stuff in my simulations. I have an experimental stress-strain data extracted from tensile tests at various temperatures and strain rates for a solder alloy. I have obtained a relatively good fit quality for ANAND model parameters (R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural Network (ANN) to find the relationships between stresses and strains, strain rates and temperatures. The ANN fit has R-sqrd of 93.1%. What I am after is that I want to conduct FEA simulations using APDL (Version 2012) to compare the response of the solder joint when exposed to thermal cycling using both material models, e.g., ANAND and ANN. Is there a direct way of having ANSYS dive into the saved ANN MATLAB file, and pull out the stress-strain relationships? I have spoken to ChatGPT and he gave me a suggestion to use USERMAT subroutine but that needs some decent programming. Another suggestion, is to use the ANN to generate stress-strain data into tabular format for different temperatures and strain rates, and then define such relationships (discrete data points) in APDL using TBDATA and TB, RATE and provided some APDL lines for it, and it looks decent but haven’t tried it yet. Does this sound doable? I mean, I know that it’s possible to define stress-strains for different temperatures but I am not sure if that’s possible for different strain rates. Is it? I would appreciate any feedback and suggestions. Best, Mohammad —- Guest Researcher University of Freiburg Germany ===================================== Mohammad A Gharaibeh, Ph.D. Associate Professor Department of Mechanical Engineering The Hashemite University P.O. Box 330127 Zarqa, 13133, Jordan Tel: +962 - 5 - 390 3333 Ext. 4771 Fax: +962 - 5 - 382 6348 =====================================

MG

Mohammad Gharaibeh

Wed, Jun 11, 2025 4:27 PM

Did this email go through?

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh mgharai1@binghamton.edu
wrote:

Dear experts,

I am trying to use some AI stuff in my simulations. I have an experimental
stress-strain data extracted from tensile tests at various temperatures and
strain rates for a solder alloy. I have obtained a relatively good fit
quality for ANAND model parameters (R-sqrd = 80.8%), and have used MATLAB
to train an Artificial Neural Network (ANN) to find the relationships
between stresses and strains, strain rates and temperatures. The ANN fit
has R-sqrd of 93.1%.

What I am after is that I want to conduct FEA simulations using APDL
(Version 2012) to compare the response of the solder joint when exposed to
thermal cycling using both material models, e.g., ANAND and ANN.

Is there a direct way of having ANSYS dive into the saved ANN MATLAB file,
and pull out the stress-strain relationships?

I have spoken to ChatGPT and he gave me a suggestion to use USERMAT
subroutine but that needs some decent programming. Another suggestion, is
to use the ANN to generate stress-strain data into tabular format for
different temperatures and strain rates, and then define such relationships
(discrete data points) in APDL using TBDATA and TB, RATE and provided some
APDL lines for it, and it looks decent but haven’t tried it yet.

Does this sound doable? I mean, I know that it’s possible to define
stress-strains for different temperatures but I am not sure if that’s
possible for different strain rates. Is it?

I would appreciate any feedback and suggestions.

Best,
Mohammad

—-
Guest Researcher
University of Freiburg
Germany

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

Did this email go through? ===================================== Mohammad A Gharaibeh, Ph.D. Associate Professor Department of Mechanical Engineering The Hashemite University P.O. Box 330127 Zarqa, 13133, Jordan Tel: +962 - 5 - 390 3333 Ext. 4771 Fax: +962 - 5 - 382 6348 ===================================== On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh <mgharai1@binghamton.edu> wrote: > > Dear experts, > > I am trying to use some AI stuff in my simulations. I have an experimental > stress-strain data extracted from tensile tests at various temperatures and > strain rates for a solder alloy. I have obtained a relatively good fit > quality for ANAND model parameters (R-sqrd = 80.8%), and have used MATLAB > to train an Artificial Neural Network (ANN) to find the relationships > between stresses and strains, strain rates and temperatures. The ANN fit > has R-sqrd of 93.1%. > > What I am after is that I want to conduct FEA simulations using APDL > (Version 2012) to compare the response of the solder joint when exposed to > thermal cycling using both material models, e.g., ANAND and ANN. > > Is there a direct way of having ANSYS dive into the saved ANN MATLAB file, > and pull out the stress-strain relationships? > > I have spoken to ChatGPT and he gave me a suggestion to use USERMAT > subroutine but that needs some decent programming. Another suggestion, is > to use the ANN to generate stress-strain data into tabular format for > different temperatures and strain rates, and then define such relationships > (discrete data points) in APDL using TBDATA and TB, RATE and provided some > APDL lines for it, and it looks decent but haven’t tried it yet. > > Does this sound doable? I mean, I know that it’s possible to define > stress-strains for different temperatures but I am not sure if that’s > possible for different strain rates. Is it? > > I would appreciate any feedback and suggestions. > > Best, > Mohammad > > —- > Guest Researcher > University of Freiburg > Germany > > ===================================== > Mohammad A Gharaibeh, Ph.D. > Associate Professor > Department of Mechanical Engineering > The Hashemite University > P.O. Box 330127 > Zarqa, 13133, Jordan > Tel: +962 - 5 - 390 3333 Ext. 4771 > Fax: +962 - 5 - 382 6348 > ===================================== > > >

MR

Matthew Ridzon, PE

Wed, Jun 11, 2025 4:31 PM

Yes, I received the original on Monday. But I had no insight to share, thus no response from me.

Best of luck my friend!

—Matt

-----Original Message-----
From: Mohammad Gharaibeh via Xansys xansys-temp@list.xansys.org
Sent: Wednesday, June 11, 2025 12:28 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Mohammad Gharaibeh mgharai1@binghamton.edu
Subject: [Xansys] Re: Material properties in APDL and ANN

Did this email go through?

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh mgharai1@binghamton.edu
wrote:

Dear experts,

I am trying to use some AI stuff in my simulations. I have an
experimental stress-strain data extracted from tensile tests at
various temperatures and strain rates for a solder alloy. I have
obtained a relatively good fit quality for ANAND model parameters
(R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural
Network (ANN) to find the relationships between stresses and strains,
strain rates and temperatures. The ANN fit has R-sqrd of 93.1%.

What I am after is that I want to conduct FEA simulations using APDL
(Version 2012) to compare the response of the solder joint when
exposed to thermal cycling using both material models, e.g., ANAND and ANN.

Is there a direct way of having ANSYS dive into the saved ANN MATLAB
file, and pull out the stress-strain relationships?

I have spoken to ChatGPT and he gave me a suggestion to use USERMAT
subroutine but that needs some decent programming. Another suggestion,
is to use the ANN to generate stress-strain data into tabular format
for different temperatures and strain rates, and then define such
relationships (discrete data points) in APDL using TBDATA and TB, RATE
and provided some APDL lines for it, and it looks decent but haven’t tried it yet.

Does this sound doable? I mean, I know that it’s possible to define
stress-strains for different temperatures but I am not sure if that’s
possible for different strain rates. Is it?

I would appreciate any feedback and suggestions.

Best,
Mohammad

—-
Guest Researcher
University of Freiburg
Germany

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.

Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list

Yes, I received the original on Monday. But I had no insight to share, thus no response from me. Best of luck my friend! —Matt -----Original Message----- From: Mohammad Gharaibeh via Xansys <xansys-temp@list.xansys.org> Sent: Wednesday, June 11, 2025 12:28 PM To: XANSYS Mailing List Home <xansys-temp@list.xansys.org> Cc: Mohammad Gharaibeh <mgharai1@binghamton.edu> Subject: [Xansys] Re: Material properties in APDL and ANN Did this email go through? ===================================== Mohammad A Gharaibeh, Ph.D. Associate Professor Department of Mechanical Engineering The Hashemite University P.O. Box 330127 Zarqa, 13133, Jordan Tel: +962 - 5 - 390 3333 Ext. 4771 Fax: +962 - 5 - 382 6348 ===================================== On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh <mgharai1@binghamton.edu> wrote: > > Dear experts, > > I am trying to use some AI stuff in my simulations. I have an > experimental stress-strain data extracted from tensile tests at > various temperatures and strain rates for a solder alloy. I have > obtained a relatively good fit quality for ANAND model parameters > (R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural > Network (ANN) to find the relationships between stresses and strains, > strain rates and temperatures. The ANN fit has R-sqrd of 93.1%. > > What I am after is that I want to conduct FEA simulations using APDL > (Version 2012) to compare the response of the solder joint when > exposed to thermal cycling using both material models, e.g., ANAND and ANN. > > Is there a direct way of having ANSYS dive into the saved ANN MATLAB > file, and pull out the stress-strain relationships? > > I have spoken to ChatGPT and he gave me a suggestion to use USERMAT > subroutine but that needs some decent programming. Another suggestion, > is to use the ANN to generate stress-strain data into tabular format > for different temperatures and strain rates, and then define such > relationships (discrete data points) in APDL using TBDATA and TB, RATE > and provided some APDL lines for it, and it looks decent but haven’t tried it yet. > > Does this sound doable? I mean, I know that it’s possible to define > stress-strains for different temperatures but I am not sure if that’s > possible for different strain rates. Is it? > > I would appreciate any feedback and suggestions. > > Best, > Mohammad > > —- > Guest Researcher > University of Freiburg > Germany > > ===================================== > Mohammad A Gharaibeh, Ph.D. > Associate Professor > Department of Mechanical Engineering > The Hashemite University > P.O. Box 330127 > Zarqa, 13133, Jordan > Tel: +962 - 5 - 390 3333 Ext. 4771 > Fax: +962 - 5 - 382 6348 > ===================================== > > > _______________________________________________ Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day. Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list

GB

Grama Bhashyam

Wed, Jun 11, 2025 6:02 PM

USERMAT will be a good option.

Regards

Grama Bhashyam
ANSYS Fellow

From: Matthew Ridzon, PE via Xansys xansys-temp@list.xansys.org
Sent: Wednesday, June 11, 2025 12:31:19 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Matthew Ridzon, PE Matt@prime-engineer.com
Subject: [Xansys] Re: Material properties in APDL and ANN

[External Sender]

Yes, I received the original on Monday. But I had no insight to share, thus no response from me.

Best of luck my friend!

—Matt

-----Original Message-----
From: Mohammad Gharaibeh via Xansys xansys-temp@list.xansys.org
Sent: Wednesday, June 11, 2025 12:28 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Mohammad Gharaibeh mgharai1@binghamton.edu
Subject: [Xansys] Re: Material properties in APDL and ANN

Did this email go through?

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh mgharai1@binghamton.edu
wrote:

Dear experts,

I am trying to use some AI stuff in my simulations. I have an
experimental stress-strain data extracted from tensile tests at
various temperatures and strain rates for a solder alloy. I have
obtained a relatively good fit quality for ANAND model parameters
(R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural
Network (ANN) to find the relationships between stresses and strains,
strain rates and temperatures. The ANN fit has R-sqrd of 93.1%.

What I am after is that I want to conduct FEA simulations using APDL
(Version 2012) to compare the response of the solder joint when
exposed to thermal cycling using both material models, e.g., ANAND and ANN.

Is there a direct way of having ANSYS dive into the saved ANN MATLAB
file, and pull out the stress-strain relationships?

I have spoken to ChatGPT and he gave me a suggestion to use USERMAT
subroutine but that needs some decent programming. Another suggestion,
is to use the ANN to generate stress-strain data into tabular format
for different temperatures and strain rates, and then define such
relationships (discrete data points) in APDL using TBDATA and TB, RATE
and provided some APDL lines for it, and it looks decent but haven’t tried it yet.

Does this sound doable? I mean, I know that it’s possible to define
stress-strains for different temperatures but I am not sure if that’s
possible for different strain rates. Is it?

I would appreciate any feedback and suggestions.

Best,
Mohammad

—-
Guest Researcher
University of Freiburg
Germany

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.

Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list

Xansys mailing list -- xansys-temp@list.xansys.org
To unsubscribe send an email to xansys-temp-leave@list.xansys.org
If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.

Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list

USERMAT will be a good option. Regards Grama Bhashyam ANSYS Fellow ________________________________ From: Matthew Ridzon, PE via Xansys <xansys-temp@list.xansys.org> Sent: Wednesday, June 11, 2025 12:31:19 PM To: XANSYS Mailing List Home <xansys-temp@list.xansys.org> Cc: Matthew Ridzon, PE <Matt@prime-engineer.com> Subject: [Xansys] Re: Material properties in APDL and ANN [External Sender] Yes, I received the original on Monday. But I had no insight to share, thus no response from me. Best of luck my friend! —Matt -----Original Message----- From: Mohammad Gharaibeh via Xansys <xansys-temp@list.xansys.org> Sent: Wednesday, June 11, 2025 12:28 PM To: XANSYS Mailing List Home <xansys-temp@list.xansys.org> Cc: Mohammad Gharaibeh <mgharai1@binghamton.edu> Subject: [Xansys] Re: Material properties in APDL and ANN Did this email go through? ===================================== Mohammad A Gharaibeh, Ph.D. Associate Professor Department of Mechanical Engineering The Hashemite University P.O. Box 330127 Zarqa, 13133, Jordan Tel: +962 - 5 - 390 3333 Ext. 4771 Fax: +962 - 5 - 382 6348 ===================================== On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh <mgharai1@binghamton.edu> wrote: > > Dear experts, > > I am trying to use some AI stuff in my simulations. I have an > experimental stress-strain data extracted from tensile tests at > various temperatures and strain rates for a solder alloy. I have > obtained a relatively good fit quality for ANAND model parameters > (R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural > Network (ANN) to find the relationships between stresses and strains, > strain rates and temperatures. The ANN fit has R-sqrd of 93.1%. > > What I am after is that I want to conduct FEA simulations using APDL > (Version 2012) to compare the response of the solder joint when > exposed to thermal cycling using both material models, e.g., ANAND and ANN. > > Is there a direct way of having ANSYS dive into the saved ANN MATLAB > file, and pull out the stress-strain relationships? > > I have spoken to ChatGPT and he gave me a suggestion to use USERMAT > subroutine but that needs some decent programming. Another suggestion, > is to use the ANN to generate stress-strain data into tabular format > for different temperatures and strain rates, and then define such > relationships (discrete data points) in APDL using TBDATA and TB, RATE > and provided some APDL lines for it, and it looks decent but haven’t tried it yet. > > Does this sound doable? I mean, I know that it’s possible to define > stress-strains for different temperatures but I am not sure if that’s > possible for different strain rates. Is it? > > I would appreciate any feedback and suggestions. > > Best, > Mohammad > > —- > Guest Researcher > University of Freiburg > Germany > > ===================================== > Mohammad A Gharaibeh, Ph.D. > Associate Professor > Department of Mechanical Engineering > The Hashemite University > P.O. Box 330127 > Zarqa, 13133, Jordan > Tel: +962 - 5 - 390 3333 Ext. 4771 > Fax: +962 - 5 - 382 6348 > ===================================== > > > _______________________________________________ Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day. Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list _______________________________________________ Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day. Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list

MG

Mohammad Gharaibeh

Wed, Jun 11, 2025 6:10 PM

Thanks, Matt and Grama, for the responses.

On that topic, I would recommend that ANSYS start thinking of ways to allow
the program to interact with MATLAB or other systems (like Python) and
allow the use for such modern material modeling techniques, i.e.,
artificial neural networks and such. Or better yet, add the capability of
creating and training AI models for material modeling.

This would be a revolutionary move for the software, in my humble opinion.

Best,
Mohammad

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

On Wed, 11 Jun 2025 at 9:04 PM Grama Bhashyam via Xansys <
xansys-temp@list.xansys.org> wrote:

USERMAT will be a good option.

Regards

Grama Bhashyam
ANSYS Fellow

From: Matthew Ridzon, PE via Xansys xansys-temp@list.xansys.org
Sent: Wednesday, June 11, 2025 12:31:19 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Matthew Ridzon, PE Matt@prime-engineer.com
Subject: [Xansys] Re: Material properties in APDL and ANN

[External Sender]

Yes, I received the original on Monday. But I had no insight to share,
thus no response from me.

Best of luck my friend!

—Matt

-----Original Message-----
From: Mohammad Gharaibeh via Xansys xansys-temp@list.xansys.org
Sent: Wednesday, June 11, 2025 12:28 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Mohammad Gharaibeh mgharai1@binghamton.edu
Subject: [Xansys] Re: Material properties in APDL and ANN

Did this email go through?

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh mgharai1@binghamton.edu
wrote:

Dear experts,

I am trying to use some AI stuff in my simulations. I have an
experimental stress-strain data extracted from tensile tests at
various temperatures and strain rates for a solder alloy. I have
obtained a relatively good fit quality for ANAND model parameters
(R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural
Network (ANN) to find the relationships between stresses and strains,
strain rates and temperatures. The ANN fit has R-sqrd of 93.1%.

What I am after is that I want to conduct FEA simulations using APDL
(Version 2012) to compare the response of the solder joint when
exposed to thermal cycling using both material models, e.g., ANAND and

ANN.

Is there a direct way of having ANSYS dive into the saved ANN MATLAB
file, and pull out the stress-strain relationships?

I have spoken to ChatGPT and he gave me a suggestion to use USERMAT
subroutine but that needs some decent programming. Another suggestion,
is to use the ANN to generate stress-strain data into tabular format
for different temperatures and strain rates, and then define such
relationships (discrete data points) in APDL using TBDATA and TB, RATE
and provided some APDL lines for it, and it looks decent but haven’t

tried it yet.

Does this sound doable? I mean, I know that it’s possible to define
stress-strains for different temperatures but I am not sure if that’s
possible for different strain rates. Is it?

I would appreciate any feedback and suggestions.

Best,
Mohammad

—-
Guest Researcher
University of Freiburg
Germany

---====
Mohammad A Gharaibeh, Ph.D.
Associate Professor
Department of Mechanical Engineering
The Hashemite University
P.O. Box 330127
Zarqa, 13133, Jordan
Tel: +962 - 5 - 390 3333 Ext. 4771
Fax: +962 - 5 - 382 6348

---====

Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an
email to xansys-temp-leave@list.xansys.org If you are receiving too many
emails from XANSYS please consider changing account settings to Digest mode
which will send a single email per day.

Please send administrative requests such as deletion from XANSYS to
xansys-mod@tynecomp.co.uk and not to the list

Xansys mailing list -- xansys-temp@list.xansys.org
To unsubscribe send an email to xansys-temp-leave@list.xansys.org
If you are receiving too many emails from XANSYS please consider changing
account settings to Digest mode which will send a single email per day.

Please send administrative requests such as deletion from XANSYS to
xansys-mod@tynecomp.co.uk and not to the list

Xansys mailing list -- xansys-temp@list.xansys.org
To unsubscribe send an email to xansys-temp-leave@list.xansys.org
If you are receiving too many emails from XANSYS please consider changing
account settings to Digest mode which will send a single email per day.

Please send administrative requests such as deletion from XANSYS to
xansys-mod@tynecomp.co.uk and not to the list

Thanks, Matt and Grama, for the responses. On that topic, I would recommend that ANSYS start thinking of ways to allow the program to interact with MATLAB or other systems (like Python) and allow the use for such modern material modeling techniques, i.e., artificial neural networks and such. Or better yet, add the capability of creating and training AI models for material modeling. This would be a revolutionary move for the software, in my humble opinion. Best, Mohammad ===================================== Mohammad A Gharaibeh, Ph.D. Associate Professor Department of Mechanical Engineering The Hashemite University P.O. Box 330127 Zarqa, 13133, Jordan Tel: +962 - 5 - 390 3333 Ext. 4771 Fax: +962 - 5 - 382 6348 ===================================== On Wed, 11 Jun 2025 at 9:04 PM Grama Bhashyam via Xansys < xansys-temp@list.xansys.org> wrote: > USERMAT will be a good option. > > Regards > > Grama Bhashyam > ANSYS Fellow > ________________________________ > From: Matthew Ridzon, PE via Xansys <xansys-temp@list.xansys.org> > Sent: Wednesday, June 11, 2025 12:31:19 PM > To: XANSYS Mailing List Home <xansys-temp@list.xansys.org> > Cc: Matthew Ridzon, PE <Matt@prime-engineer.com> > Subject: [Xansys] Re: Material properties in APDL and ANN > > [External Sender] > > Yes, I received the original on Monday. But I had no insight to share, > thus no response from me. > > Best of luck my friend! > > —Matt > > -----Original Message----- > From: Mohammad Gharaibeh via Xansys <xansys-temp@list.xansys.org> > Sent: Wednesday, June 11, 2025 12:28 PM > To: XANSYS Mailing List Home <xansys-temp@list.xansys.org> > Cc: Mohammad Gharaibeh <mgharai1@binghamton.edu> > Subject: [Xansys] Re: Material properties in APDL and ANN > > Did this email go through? > > ===================================== > Mohammad A Gharaibeh, Ph.D. > Associate Professor > Department of Mechanical Engineering > The Hashemite University > P.O. Box 330127 > Zarqa, 13133, Jordan > Tel: +962 - 5 - 390 3333 Ext. 4771 > Fax: +962 - 5 - 382 6348 > ===================================== > > > > > On Mon, 9 Jun 2025 at 2:39 PM Mohammad Gharaibeh <mgharai1@binghamton.edu> > wrote: > > > > > Dear experts, > > > > I am trying to use some AI stuff in my simulations. I have an > > experimental stress-strain data extracted from tensile tests at > > various temperatures and strain rates for a solder alloy. I have > > obtained a relatively good fit quality for ANAND model parameters > > (R-sqrd = 80.8%), and have used MATLAB to train an Artificial Neural > > Network (ANN) to find the relationships between stresses and strains, > > strain rates and temperatures. The ANN fit has R-sqrd of 93.1%. > > > > What I am after is that I want to conduct FEA simulations using APDL > > (Version 2012) to compare the response of the solder joint when > > exposed to thermal cycling using both material models, e.g., ANAND and > ANN. > > > > Is there a direct way of having ANSYS dive into the saved ANN MATLAB > > file, and pull out the stress-strain relationships? > > > > I have spoken to ChatGPT and he gave me a suggestion to use USERMAT > > subroutine but that needs some decent programming. Another suggestion, > > is to use the ANN to generate stress-strain data into tabular format > > for different temperatures and strain rates, and then define such > > relationships (discrete data points) in APDL using TBDATA and TB, RATE > > and provided some APDL lines for it, and it looks decent but haven’t > tried it yet. > > > > Does this sound doable? I mean, I know that it’s possible to define > > stress-strains for different temperatures but I am not sure if that’s > > possible for different strain rates. Is it? > > > > I would appreciate any feedback and suggestions. > > > > Best, > > Mohammad > > > > —- > > Guest Researcher > > University of Freiburg > > Germany > > > > ===================================== > > Mohammad A Gharaibeh, Ph.D. > > Associate Professor > > Department of Mechanical Engineering > > The Hashemite University > > P.O. Box 330127 > > Zarqa, 13133, Jordan > > Tel: +962 - 5 - 390 3333 Ext. 4771 > > Fax: +962 - 5 - 382 6348 > > ===================================== > > > > > > > _______________________________________________ > Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an > email to xansys-temp-leave@list.xansys.org If you are receiving too many > emails from XANSYS please consider changing account settings to Digest mode > which will send a single email per day. > > Please send administrative requests such as deletion from XANSYS to > xansys-mod@tynecomp.co.uk and not to the list > _______________________________________________ > Xansys mailing list -- xansys-temp@list.xansys.org > To unsubscribe send an email to xansys-temp-leave@list.xansys.org > If you are receiving too many emails from XANSYS please consider changing > account settings to Digest mode which will send a single email per day. > > Please send administrative requests such as deletion from XANSYS to > xansys-mod@tynecomp.co.uk and not to the list > _______________________________________________ > Xansys mailing list -- xansys-temp@list.xansys.org > To unsubscribe send an email to xansys-temp-leave@list.xansys.org > If you are receiving too many emails from XANSYS please consider changing > account settings to Digest mode which will send a single email per day. > > Please send administrative requests such as deletion from XANSYS to > xansys-mod@tynecomp.co.uk and not to the list

GB

Grama Bhashyam

Wed, Jun 11, 2025 6:15 PM

Yes, we already have pyMAPDL, PyAnsys and many others. We have seen an example of USERMAT in python but too slow. Please contact your support team for more on Python interfaces

Regards

Grama Bhashyam
ANSYS Fellow

From: Mohammad Gharaibeh via Xansys xansys-temp@list.xansys.org
Sent: Wednesday, June 11, 2025 2:10:34 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Mohammad Gharaibeh mgharai1@binghamton.edu
Subject: [Xansys] Re: [External Email] Re: Material properties in APDL and ANN