Empathy List Archives

EP

Eisentraeger, Peter

Wed, Sep 1, 2021 12:44 PM

Hello all,

I am doing a transient analysis where a block of steel may lift off or not. Most likely is does not. The block has dimensions 0.4mx0.8m and rests on a track also made off steel. I am interested does it lift off and how high is the acceleration meters away from that point. My simple model is a spring node-2-node conta178 where I give the stiffness and the TOLN value like
r,what_real,-1e12,,2
rmodif,what_real,9,-10e-6 ! TOLN
The stiffness is the highest value in the whole FEM. I know the higher the stiffness and the smaller the TOLN the higher is the acceleration I am looking for.

Does anybody have experience how high I should set the stiffness or how tight I should set the tolerance?

Mit freundlichen Grüßen / Kind regards

i.A. Peter Eisenträger
Senior CAE Engineer

OHB Digital Connect GmbH
Weberstraße 21
55130 Mainz / Germany

Tel. +49-6131-2777-106
Fax. +49-6131-2777-205
eMail: Peter.Eisentraeger@ohb.de mailto:Peter.Eisentraeger@ohb.de
Internet: www.ohb-digital.de http://www.ohb-digital.de/

Geschäftsführer: Dennis Winkelmann, Dr. Dieter Birreck
Sitz der Gesellschaft: Bremen
Registergericht: Amtsgericht Bremen, HRB 20042

[https://www.ohb.de/emailbanner.png]

Hello all, I am doing a transient analysis where a block of steel may lift off or not. Most likely is does not. The block has dimensions 0.4mx0.8m and rests on a track also made off steel. I am interested does it lift off and how high is the acceleration meters away from that point. My simple model is a spring node-2-node conta178 where I give the stiffness and the TOLN value like r,what_real,-1e12,,2 rmodif,what_real,9,-10e-6 ! TOLN The stiffness is the highest value in the whole FEM. I know the higher the stiffness and the smaller the TOLN the higher is the acceleration I am looking for. Does anybody have experience how high I should set the stiffness or how tight I should set the tolerance? Mit freundlichen Grüßen / Kind regards i.A. Peter Eisenträger Senior CAE Engineer ------------------------------------------- OHB Digital Connect GmbH Weberstraße 21 55130 Mainz / Germany Tel. +49-6131-2777-106 Fax. +49-6131-2777-205 eMail: Peter.Eisentraeger@ohb.de<mailto:Peter.Eisentraeger@ohb.de> Internet: www.ohb-digital.de<http://www.ohb-digital.de/> Geschäftsführer: Dennis Winkelmann, Dr. Dieter Birreck Sitz der Gesellschaft: Bremen Registergericht: Amtsgericht Bremen, HRB 20042 ------------------------------------------- [https://www.ohb.de/emailbanner.png]

WG

Westwater, Greg [AUTOSOL/FCV/MTWN]

Wed, Sep 1, 2021 2:01 PM

You might try using a pre-stress modal analysis and tuning the contact stiffness to match test data (if available). Years ago one of our analysts determined that tunning the normal contact stiffness allowed us to match test data. I dug into it and learned that tuning the stiffness is an accepted approach and helps account for physics at the interface not being captured in the numerical model. It sounds like you are manually defining the stiffness...that may be helpful. If you only adjust the stiffness factor you have to be very careful with multiple settings...left to it's own devices the ANSYS algorithms will sometimes give you unexpected interactions. In general I find ANSYS tends to overpredict contact stiffness...this minimizes penetration error for most structural analysis situations, but ends up over predicting natural frequency, at least for the types of structures we work with.

I'm not sure if my suggestion is valid or helpful, but my thought is that if you've matched the stiffness appropriately for a modal response, then the same stiffness may be appropriate for your transient analysis.

Gregory Westwater

-----Original Message-----
From: Eisentraeger, Peter peter.eisentraeger@ohb.de
Sent: Wednesday, September 1, 2021 7:44 AM
To: xansys-temp@list.xansys.org
Subject: [Xansys] Stiffness of steel-steel contact

Hello all,

I am doing a transient analysis where a block of steel may lift off or not. Most likely is does not. The block has dimensions 0.4mx0.8m and rests on a track also made off steel. I am interested does it lift off and how high is the acceleration meters away from that point. My simple model is a spring node-2-node conta178 where I give the stiffness and the TOLN value like
r,what_real,-1e12,,2
rmodif,what_real,9,-10e-6 ! TOLN The stiffness is the highest value in the whole FEM. I know the higher the stiffness and the smaller the TOLN the higher is the acceleration I am looking for.

Does anybody have experience how high I should set the stiffness or how tight I should set the tolerance?

Mit freundlichen Grüßen / Kind regards

i.A. Peter Eisenträger
Senior CAE Engineer

OHB Digital Connect GmbH
Weberstraße 21
55130 Mainz / Germany

Tel. +49-6131-2777-106
Fax. +49-6131-2777-205
eMail: Peter.Eisentraeger@ohb.de mailto:Peter.Eisentraeger@ohb.de
Internet: https://urldefense.proofpoint.com/v2/url?u=http-3A__www.ohb-2Ddigital.de&d=DwIFAw&c=jOURTkCZzT8tVB5xPEYIm3YJGoxoTaQsQPzPKJGaWbo&r=SI6AZzvnz60GdYZHf-785aZ7BG-N1Wp80WNx-PjYGdo&m=zE4qLdcs1_BCzvBNozu8a04rIahgsVxWhkvisyXRklY&s=WGRmz7q_uT7Ogs6LX1np8NRqnPAyupQDQgCYaFnjhT8&e= <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.ohb-2Ddigital.de_&d=DwIFAw&c=jOURTkCZzT8tVB5xPEYIm3YJGoxoTaQsQPzPKJGaWbo&r=SI6AZzvnz60GdYZHf-785aZ7BG-N1Wp80WNx-PjYGdo&m=zE4qLdcs1_BCzvBNozu8a04rIahgsVxWhkvisyXRklY&s=V_LTm_Ae1BkKvXA13BbXHNzH4QDGR9SIccFKlRhH0ZM&e= >

Geschäftsführer: Dennis Winkelmann, Dr. Dieter Birreck Sitz der Gesellschaft: Bremen
Registergericht: Amtsgericht Bremen, HRB 20042

[https://urldefense.proofpoint.com/v2/url?u=https-3A__www.ohb.de_emailbanner.png&d=DwIFAw&c=jOURTkCZzT8tVB5xPEYIm3YJGoxoTaQsQPzPKJGaWbo&r=SI6AZzvnz60GdYZHf-785aZ7BG-N1Wp80WNx-PjYGdo&m=zE4qLdcs1_BCzvBNozu8a04rIahgsVxWhkvisyXRklY&s=a_hP2sh6WRxCVSgr7dWZP5r7JsAKW2gGMhm0onSrqrw&e= ]

You might try using a pre-stress modal analysis and tuning the contact stiffness to match test data (if available). Years ago one of our analysts determined that tunning the normal contact stiffness allowed us to match test data. I dug into it and learned that tuning the stiffness is an accepted approach and helps account for physics at the interface not being captured in the numerical model. It sounds like you are manually defining the stiffness...that may be helpful. If you only adjust the stiffness factor you have to be very careful with multiple settings...left to it's own devices the ANSYS algorithms will sometimes give you unexpected interactions. In general I find ANSYS tends to overpredict contact stiffness...this minimizes penetration error for most structural analysis situations, but ends up over predicting natural frequency, at least for the types of structures we work with. I'm not sure if my suggestion is valid or helpful, but my thought is that if you've matched the stiffness appropriately for a modal response, then the same stiffness may be appropriate for your transient analysis. Gregory Westwater Gregory Westwater | Senior Engineering Specialist | Flow Controls Emerson Automation Solutions | 301 S. 1st Ave. | Marshalltown | Iowa | 50158 | USA Greg.Westwater@Emerson.com -----Original Message----- From: Eisentraeger, Peter <peter.eisentraeger@ohb.de> Sent: Wednesday, September 1, 2021 7:44 AM To: xansys-temp@list.xansys.org Subject: [Xansys] Stiffness of steel-steel contact Hello all, I am doing a transient analysis where a block of steel may lift off or not. Most likely is does not. The block has dimensions 0.4mx0.8m and rests on a track also made off steel. I am interested does it lift off and how high is the acceleration meters away from that point. My simple model is a spring node-2-node conta178 where I give the stiffness and the TOLN value like r,what_real,-1e12,,2 rmodif,what_real,9,-10e-6 ! TOLN The stiffness is the highest value in the whole FEM. I know the higher the stiffness and the smaller the TOLN the higher is the acceleration I am looking for. Does anybody have experience how high I should set the stiffness or how tight I should set the tolerance? Mit freundlichen Grüßen / Kind regards i.A. Peter Eisenträger Senior CAE Engineer ------------------------------------------- OHB Digital Connect GmbH Weberstraße 21 55130 Mainz / Germany Tel. +49-6131-2777-106 Fax. +49-6131-2777-205 eMail: Peter.Eisentraeger@ohb.de<mailto:Peter.Eisentraeger@ohb.de> Internet: https://urldefense.proofpoint.com/v2/url?u=http-3A__www.ohb-2Ddigital.de&d=DwIFAw&c=jOURTkCZzT8tVB5xPEYIm3YJGoxoTaQsQPzPKJGaWbo&r=SI6AZzvnz60GdYZHf-785aZ7BG-N1Wp80WNx-PjYGdo&m=zE4qLdcs1_BCzvBNozu8a04rIahgsVxWhkvisyXRklY&s=WGRmz7q_uT7Ogs6LX1np8NRqnPAyupQDQgCYaFnjhT8&e= <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.ohb-2Ddigital.de_&d=DwIFAw&c=jOURTkCZzT8tVB5xPEYIm3YJGoxoTaQsQPzPKJGaWbo&r=SI6AZzvnz60GdYZHf-785aZ7BG-N1Wp80WNx-PjYGdo&m=zE4qLdcs1_BCzvBNozu8a04rIahgsVxWhkvisyXRklY&s=V_LTm_Ae1BkKvXA13BbXHNzH4QDGR9SIccFKlRhH0ZM&e= > Geschäftsführer: Dennis Winkelmann, Dr. Dieter Birreck Sitz der Gesellschaft: Bremen Registergericht: Amtsgericht Bremen, HRB 20042 ------------------------------------------- [https://urldefense.proofpoint.com/v2/url?u=https-3A__www.ohb.de_emailbanner.png&d=DwIFAw&c=jOURTkCZzT8tVB5xPEYIm3YJGoxoTaQsQPzPKJGaWbo&r=SI6AZzvnz60GdYZHf-785aZ7BG-N1Wp80WNx-PjYGdo&m=zE4qLdcs1_BCzvBNozu8a04rIahgsVxWhkvisyXRklY&s=a_hP2sh6WRxCVSgr7dWZP5r7JsAKW2gGMhm0onSrqrw&e= ]

CW

Christopher Wright

Wed, Sep 1, 2021 7:53 PM

On Sep 1, 2021, at 7:44 AM, Eisentraeger, Peter peter.eisentraeger@ohb.de wrote:

Does anybody have experience how high I should set the stiffness or how tight I should set the tolerance?

If it were my problem (I've had a few like it)I'd run a very simple linear analysis and check to see whether it actually lifts. If it doesn't lift or almost doesn't lift, you have your solution without a lot of work and tweaking around with parameter studies.You also have a debugged model you can easily fiddle to add a contact element if that becomes necessary. By rule of thumb the gap stiffness should be 5-10 times the local model stiffness; tolerances should conform to the local design tolerances.

If you get significant tensile contact forces, then of course you may feel the need to add a contact element and go whole hog. But at least you tried.

Christopher Wright P.E. (ret'd) |"They couldn't hit an elephant at
chrisw@skypoint.com | this distance" (last words of Gen.
| John Sedgwick, Spotsylvania (1864)
http://www.skypoint.com/members/chrisw/

> On Sep 1, 2021, at 7:44 AM, Eisentraeger, Peter <peter.eisentraeger@ohb.de> wrote: > > Does anybody have experience how high I should set the stiffness or how tight I should set the tolerance? If it were my problem (I've had a few like it)I'd run a very simple linear analysis and check to see whether it actually lifts. If it doesn't lift or almost doesn't lift, you have your solution without a lot of work and tweaking around with parameter studies.You also have a debugged model you can easily fiddle to add a contact element if that becomes necessary. By rule of thumb the gap stiffness should be 5-10 times the local model stiffness; tolerances should conform to the local design tolerances. If you get significant tensile contact forces, then of course you may feel the need to add a contact element and go whole hog. But at least you tried. Christopher Wright P.E. (ret'd) |"They couldn't hit an elephant at chrisw@skypoint.com | this distance" (last words of Gen. | John Sedgwick, Spotsylvania (1864) http://www.skypoint.com/members/chrisw/