Hello Mr. Barkey, Thank You very much for this very informative and knowledgeable video. I am using Ramberg-Osgood stress -strain curve and Eq. 13.61 from "Mechanical Behavior of Materials" Norman E. Dowling (Fourth Edition) for the calculations. The analytical approach is similar to the described in your lecture. I could not find your next lecture of this topic. What is the link of your next lecture/video, where the stresses are calculated from the estimated strains, calculated from Ramberg-Osgood equation, in your current lecture/video. I need to convert my FEA stresses (> Yield Stresses) from linear elastic FE analysis to elastic-plastic stresses. Thanks for your reply and guidance.
I am not exactly sure that you are looking for but you might find it in either my Plasticity or my Fatigue Analysis playlists. If it is not there, then I likely do not have a video for it. Happy studies!
I am using a formula to check the results getting from FEM analysis. That formula contains the R-O model constants alpha and n. So I am bound to model this material by R_O model. Is there any material model in ABAQUS by which I can model this AISI 1030 flow curve more accurately.
@@mbarkey.mechanics In the flow curve after yielding just started, the stress value remains constant for increment in strain upto some label. Then the curve starts moving up again. How to tackle that part. I mean to say for a given yield stress value I will not have zero plastic strain. But in the ABAQUS input the plastic strain must be zero for yield stress. So how to tackle that problem sir.
You might try to give it a very small slope, e.g. if yield stress = 200 MPa, use 200, 0; 201, 0.001, 202; 0.002 etc. and then use the rest of the curve as you normally would.
I have a true stress strain curve for AISI-1030 . I know the yield stress and young's modulus of the material. Can i find the Ramberg-Osgood coefficients alpha and n by performing a curve fitting on the whole set of data? I am using this form of R-O model.(strain/yield stress)=(stress/yield stress)+alp*(stress/yield stress)^(n+1) .stress and strain are true values not engineering.
I really think you should look to make sure that equation is OK. Some people will normalize stress by a yield stress or reference stress value, but I'm not sure it makes sense to also normalize strain by yield stress because the units on those terms are not consistent. But maybe you meant to type yield strain on the left side, and so I will continue with that assumption. First of all, the R-O equation is a curve fit, and you can always make your own curve fit or equation. Now, the equation that you present does not seem to have much relation to the parameters E, K, n in the R-O equation that I have presented. The R-O equation has no yield stress in it--for numerical reasons, you can partition it into "insignificant" and "significant" plasticity, but the boundary value for this, although called a yield stress, has little relation to what you might ordinarily consider a yield stress. Your AISI-1030 steel looks like it is low carbon, and neither your equation nor the R-O will pick up any double yield point behavior if it exists for your material. If the elastic strain is really related by linear Hooke's law, then I am not seeing that in your equation either--and the capturing the power law behavior of plastic strains is what the R-O equation is all about. (And not all metals are modelled very well as a power law material.)
Good evening, professor! Around min. 14, when you start working on Excel, how did you calculate the true strain and true stress that appear at the beginning? By converting the engineering strain and stress that we obtain in a tension test with: epsilon_t = ln (1+epsilon_e) sigma_t = sigma_e (1+ epsilon_e) ? Thank you!
Professor, I'm really sorry for bothering you, but I would really appreciate some help, please. I did a tension test on a flat specimen. My data are displacement and force. I need to obtain: yield stress, maximum stress/ultimate stress, and rupture stress, so I did: 1. engineering curve with my data and also by doing the preload correction; 2. calculated yield stress with that curve (or should it be with the true curve?) and the 0.2% offset method; 3. calculated true strain and true stress with epsilon_t = ln (1+epsilon_e) + sigma_t = sigma_e (1+ epsilon_e); 4. corrected my true curve with the RO equation like this video (I plotted that curve with all of my points, not just to the maximum point, but the curve stops at that point, how do I extend it until the fracture in order to get my true curve corrected and completed? (that question is because I want to plot the engineering curve and the true curve in the same graph so that it is possible to see the difference). I thought that the RO method would fix the true curve, calculated with epsilon_t = ln (1+epsilon_e) and sigma_t = sigma_e (1+ epsilon_e), because of the triaxial state of stress, so the result would be a "straight line" after necking and not the same curve trend of engineering curve and the true curve. is this wrong? So my questions are: 1. after applying the RO equation, how do I extend the curve until the fracture in order to get my true curve corrected and completed? (like I questioned above) 2. how do I obtain my maximum stress? by finding the maximum true stress ( sigma_e (1+ epsilon_e) ) in my points? or in the engineering curve? 3. how do I obtain my rupture stress? is it just the last point of my true curve? or my engineering curve? I really appreciate your time and help!
Furthermore, my true curve ( sigma_e (1+ epsilon_e) ) and the true curve with RO are coincident up to the maximum point (as expected), where the true curve of RO ends. So why does the RO equation apply?
@@anaritalourenco9472 Hi Ana Rita--this sounds like it may be a homework problem? or maybe you do not have the background yet for this type of analysis? Please clarify what you mean by displacement--is this cross head displacement? Did you use an extensometer during your test? (A stress-strain curve test needs to be done with an extensometer of some sort. This video will explain why it matters: ua-cam.com/video/jy2VB8GzGig/v-deo.html ) Regarding the RO equation, it is just a curve fit of the data. It does not fix or correct anything. Conversion of engineering stress-strain data to true stress strain data by the equations you mention are not valid after necking. If you want something during and after necking, look for the term "Bridgeman correction".
Excellent lecture!, very discriptive and easy to understand. The voice and pace of the lecture is easy to understand.
Thanks! Happy studies!
Hello Mr. Barkey,
Thank You very much for this very informative and knowledgeable video.
I am using Ramberg-Osgood stress -strain curve and Eq. 13.61 from "Mechanical Behavior of Materials" Norman E. Dowling (Fourth Edition) for the calculations.
The analytical approach is similar to the described in your lecture. I could not find your next lecture of this topic.
What is the link of your next lecture/video, where the stresses are calculated from the estimated strains, calculated from Ramberg-Osgood equation, in your current lecture/video.
I need to convert my FEA stresses (> Yield Stresses) from linear elastic FE analysis to elastic-plastic stresses.
Thanks for your reply and guidance.
I am not exactly sure that you are looking for but you might find it in either my Plasticity or my Fatigue Analysis playlists. If it is not there, then I likely do not have a video for it. Happy studies!
You can double click on the bottom right corner of the cell which is you want to copy. And also you can use F4 to fix cell ($)
Thank you for your informative videos.
I am using a formula to check the results getting from FEM analysis. That formula contains the R-O model constants alpha and n. So I am bound to model this material by R_O model. Is there any material model in ABAQUS by which I can model this AISI 1030 flow curve more accurately.
If that formula does not always work, you can always use direct input of stress vs plastic strain.
@@mbarkey.mechanics In the flow curve after yielding just started, the stress value remains constant for increment in strain upto some label. Then the curve starts moving up again. How to tackle that part. I mean to say for a given yield stress value I will not have zero plastic strain. But in the ABAQUS input the plastic strain must be zero for yield stress. So how to tackle that problem sir.
You might try to give it a very small slope, e.g. if yield stress = 200 MPa, use 200, 0; 201, 0.001, 202; 0.002 etc. and then use the rest of the curve as you normally would.
This video may be useful: ua-cam.com/video/-Vnk3KYomV8/v-deo.html&list=PLaDWa6xI4zefL69_xDs-OoS8pECVpAo3F&index=12
I have a true stress strain curve for AISI-1030 . I know the yield stress and young's modulus of the material. Can i find the Ramberg-Osgood coefficients alpha and n by performing a curve fitting on the whole set of data? I am using this form of R-O model.(strain/yield stress)=(stress/yield stress)+alp*(stress/yield stress)^(n+1) .stress and strain are true values not engineering.
I really think you should look to make sure that equation is OK. Some people will normalize stress by a yield stress or reference stress value, but I'm not sure it makes sense to also normalize strain by yield stress because the units on those terms are not consistent. But maybe you meant to type yield strain on the left side, and so I will continue with that assumption.
First of all, the R-O equation is a curve fit, and you can always make your own curve fit or equation. Now, the equation that you present does not seem to have much relation to the parameters E, K, n in the R-O equation that I have presented. The R-O equation has no yield stress in it--for numerical reasons, you can partition it into "insignificant" and "significant" plasticity, but the boundary value for this, although called a yield stress, has little relation to what you might ordinarily consider a yield stress. Your AISI-1030 steel looks like it is low carbon, and neither your equation nor the R-O will pick up any double yield point behavior if it exists for your material. If the elastic strain is really related by linear Hooke's law, then I am not seeing that in your equation either--and the capturing the power law behavior of plastic strains is what the R-O equation is all about. (And not all metals are modelled very well as a power law material.)
Good evening, professor! Around min. 14, when you start working on Excel, how did you calculate the true strain and true stress that appear at the beginning? By converting the engineering strain and stress that we obtain in a tension test with:
epsilon_t = ln (1+epsilon_e)
sigma_t = sigma_e (1+ epsilon_e) ?
Thank you!
Yes, please refer to about 34:00 on this video: ua-cam.com/video/GR6CpWxT5og/v-deo.html
@@mbarkey.mechanics Thank you, professor!!
Professor, I'm really sorry for bothering you, but I would really appreciate some help, please.
I did a tension test on a flat specimen. My data are displacement and force.
I need to obtain: yield stress, maximum stress/ultimate stress, and rupture stress, so I did:
1. engineering curve with my data and also by doing the preload correction;
2. calculated yield stress with that curve (or should it be with the true curve?) and the 0.2% offset method;
3. calculated true strain and true stress with epsilon_t = ln (1+epsilon_e) + sigma_t = sigma_e (1+ epsilon_e);
4. corrected my true curve with the RO equation like this video (I plotted that curve with all of my points, not just to the maximum point, but the curve stops at that point, how do I extend it until the fracture in order to get my true curve corrected and completed? (that question is because I want to plot the engineering curve and the true curve in the same graph so that it is possible to see the difference). I thought that the RO method would fix the true curve, calculated with epsilon_t = ln (1+epsilon_e) and sigma_t = sigma_e (1+ epsilon_e), because of the triaxial state of stress, so the result would be a "straight line" after necking and not the same curve trend of engineering curve and the true curve. is this wrong?
So my questions are:
1. after applying the RO equation, how do I extend the curve until the fracture in order to get my true curve corrected and completed? (like I questioned above)
2. how do I obtain my maximum stress? by finding the maximum true stress ( sigma_e (1+ epsilon_e) ) in my points? or in the engineering curve?
3. how do I obtain my rupture stress? is it just the last point of my true curve? or my engineering curve?
I really appreciate your time and help!
Furthermore, my true curve ( sigma_e (1+ epsilon_e) ) and the true curve with RO are coincident up to the maximum point (as expected), where the true curve of RO ends. So why does the RO equation apply?
@@anaritalourenco9472 Hi Ana Rita--this sounds like it may be a homework problem? or maybe you do not have the background yet for this type of analysis? Please clarify what you mean by displacement--is this cross head displacement? Did you use an extensometer during your test? (A stress-strain curve test needs to be done with an extensometer of some sort. This video will explain why it matters: ua-cam.com/video/jy2VB8GzGig/v-deo.html ) Regarding the RO equation, it is just a curve fit of the data. It does not fix or correct anything. Conversion of engineering stress-strain data to true stress strain data by the equations you mention are not valid after necking. If you want something during and after necking, look for the term "Bridgeman correction".