User login


You are here

Concept of Yield Point

I'm an undergraduate who has just completed the Mechanics of materials Course.I'm still confused about what happens at the yield point.In the stress strain curve how does the strain increase accompanied by a decrease in stress?can u please throw light on this small confusion of mine?


Do mechanicians need software or can they do without learning them ?
What softwares must an aspiring mechanician learn?

Temesgen Markos's picture

Hi Vijay,

I would split the software literacy of engineers into two: programming langauges and application software. I really can't say which programming langauge one has to learn; if you learn one of C++, Fortran or C# properly that would be enough. I guess you are an undergraduate student. Most probably your university will choose which programming langauge you are going to learn. Learn that very well and it's quite easy to shift to another language if needed. If you are mathematically inclined, it would be good if you also learn Matlab. It is very widely used for technical computing. Maple or Mathematica can be useful if you have to do symbolic calculations. Matlab understands Maple commands.

The other category is application software. Graphics packages (such as AutoCAD), project management tools, mechanism analysis packages and the like which are used for carrying out a specific task fall under this class. You are more likely to uses these ones as a professional than as a student. The specific brand used depends on company and geographic preferences. IMO AutoCAD is very widely used among Mechanical, and Civil engineers. Probably your school will introduce you to this software in you engineering drawing/graphics courses.

Finally, it would be good to also learn a finite element software. In your senior undergraduate or beginning graduate years you will most probably have a course in finite element methods and will be introduced to one or more FEM software. 

I think I went ahead of myself talking about software choice with out saying if at all they are a must. I would say it is possible to do with out but this will be extremely difficult.

Thanks a lot for your guidance

Arun Prakash's picture

Do mechanicians need software or can they do without learning them ?

Yes  for sure. Now by software, i do not mean just commercial softwares. It could be self programmed ones. As David suggests, a good mechanician should be able to solve problems without a computer. But then one is restricted to simple problems. It is to help us solve complicated problems of day-to-day nature, that we actually use softwares. Having said that, one must note that all softwares have a theory behind them. The theory must and should be well understood by the user, if he/she aspires to analyze and solve the problem more efficiently. Infact, with the knowledge of theory at hand, you should be,theoretically, able to write your own software.

 As for the important softwares that mechanicians need to learn - I would say one programming language atleast. It does not matter which one. But I would seriously suggest an object oriented programming language like C++. Fortran is not a bad choice either as a lot of commercial FE packages have user defined subroutines in Fortran.

In response to your two posts:

1. The yield point signifies when the material starts to give.  Consider an object made of some elastically deformable material (e.g. rubber).  Changing the shape of the object results in stresses that resist the deviation from the original shape.  As long as the material has not reached its yield point, returning the object to its original shape will remove the stresses.  If, however, the material has been deformed beyond its yield point, then this is no longer true; the unstressed shape of the object has changed, and in many cases, further changes in shape are met with less resistance.

2.  Just as a good mathematician should be able to figure out how to solve math problems without a calculator, a good mechanician should be able to figure out how to solve mechanics problems without a computer.  That said, any program a mechanician uses to help solve problems should effectively carry out what they would have done with nothing more than pencils & paper (and a considerable amount of time).  Whether or not commercially available software actually does this is another question.

Thanks a lot for throwing some light on these topics...I'm more clear abt it now...

Arun Prakash's picture


 The yield point is a point on the stress strain surve after which the material yeilds!. In other words, the yield point signifies the end of elastic region for the material. Crossing this point means that a certain amount of energy is lost through dissipation and cannot be recovered. What is recovered after unloading is only the elastic part of the deformation. The inelastic behavior can be viscous, plastic or visco-plastic in nature.

 When you talk of decrease in stress due to increase in strain, i assume you are mainly referring to the nominal(engineering stress)/nominal strain diagram. In this case one measures stress with respect to the original c/s area. However, in case of a true stress/true strain curve, you may not see such a thing. Of course different materials show different characteristics. Aluminium for e.g. does not show any such decrease in stress. 

The concept of Y.P can be extended to a complete surface called the Yield surface which helps mechanicians analyze the material characteristics better.


Thanks for the wonderful explanation abt yield point n ur guidance abt software packages.i'm into it now...


You won't get these ideas fully well by going through just one course in the Mechanics of Materials alone. You will also need other courses. Understanding and maturation of some concepts does take some finite quantum of time---they cannot be hurried too much. Further, what helps in such maturation is repeated thinking about the same issue but from various angles, e.g., as given by different courses or different text-book covering different primary subject matter.

To give you some specific help in this regard, I would recommend the relevant topics from the following books, roughly in the order given:

(i) Callister (Materials Science) or van Vlack (Materials Science), and then, Avner (Physical Metallurgy).

(ii) If you are not yet satisfied and still need some further detailed explanations, see Reed-Hill (Physical Metallurgy) which very possibly might be listed in the "Reference Books" section of your syllabus, but forms a very easy reading book despite its advanced content. Courtney (Mechanical Behavior) would be slightly more specialized, but worth flipping through. Dieter (Mechanical Metallurgy) is good but would be even harder to read at this stage of your preparation---primarily because the relevant descriptions are given in several different places in that book; Dieter assumes that you already had a couple of other courses beforehand.

BTW, note a few things:

(i) Yield point, as a clearly demarcated point on the engg. stress--engg. strain graph, a point after which the stress momentarily drops and the stress-strain curve experiences a very small "flat" patch with jaggedness in it, before stress begins to rise again and work-hardening begins, is a behavior found only in certain steels. Only steels show this kind of a pronounced yield point. Nonferrous materials like copper and aluminium do not show this kind of clearly demarcated yield point. (The reason steels show the behaviour has got to do with the beginning of the breaking away of the hitherto pinned down dislocations from the carbon atom "atmosphere".)

(ii) There is another point on the stress-strain curve where stress is once again seen to drop with increasing stress, but such a behavior is shown only in the engg. stress vs. engg. strain plot, not the true stress vs true strain plot (as Arun mentioned above). Such a behaviour is shown when you go beyond the point of UTS (ultimate tensile strength)---i.e. the highest point on the engg. stress-strain curve. The important phenomenon here is necking, not yielding.

(iii) The word "yielding" literally means "giving." The usage of the word becomes clear once you understand its historical context and the intended metaphor. The metaphor being used here is that of a duel---a physical fight between two fighters, fought till the very end (presumably, of only one of the two).

Actually, the amount of strain *before* the yield point is extremely small in terms of its absolute magnitude. It is unnoticeably small, in fact, unless you are using very sensitive instrumentation.

In comparison, the amount of strain *after* the yield point is quite large. It also happens to be a plastic in nature---i.e. it is a permanent deformation.

The intended allegory here is to imagine as if the test specimen were putting up a fight of some sort with the applied load---a fight which it would eventually lose.

Until the yield point, in this allegory, the material is seen to be fighting back equally well, without "giving in even an inch of the ground" because it does not seem to show any "permanent set."(Really speaking, there *is* a small elastic strain even before the yield point, but its magnitude is so small that it need not be taken very seriously in an allegory.)

After the yield point, the material does begin to show the inescapable signs of very larege and permanently suffered changes to itself, as if it were now losing the battle, or "giving in," to the incessantly applied force. Hence the name: the "giving-in point" or the "yield point".

Notice here that it was Robert Hooke who first plotted the stress-strain curve in the 17th century, and for the next 150 years or so, up to the early 19th century, duels formed a "normal" social practice to end disputes, esp. in Europe and America. In fact, the world lost an extraordinarily talented mathematician by name "Galois" to a duel---when he had not even completed 21 years of his age. ...

I am not sure if it was Robert Hooke who named the point of the onset of permanent deformation as Yield Point, but whoever named it must have found an emotionally very evocative (and arguably, therefore, a very easily recognized) circumstance to name the phenomenon with. Something like, in today's world, naming the shape that a drop of milk assumes after falling into a glass of clear water, as: "mushroom cloud." Instant recognition would be almost guaranteed with this kind of an evocative name!

The same person could be playing multiple roles: as a mechanician and as a software builder---programmer, designer, etc.   

As far as the role as a mechanician goes, software is primarily only a *tool* But as a "soft" tool, it has extraordinary range of applications, and therefore, power, as compared to the "hard" tools like machine tools. (You can use a physical model of a dam only for a single dam and only for that purpose---fluid dynamical simulation of a limited range. For a physical model of an aeroplane, you have to go wind tunnel---another setup, another model. In contrast, the same supercomputer can easily act as both within a matter of hours for you.) 

However, as a soft tool, software necessarily makes use of *mathematical* models of certain pre-isolated physical phenomena. This, essentially, sets the fundamental limits for its utility. You can only use software where the basic mechanical model and physical laws such as the constitutional law have already been precisely determined.

Yet, a further role for software---arising out of its ease of use and power these days---is not only as a tool for modeling and simulation but also as an enabling tool for discovery. 

At the undergraduate stage of one's education as a non-computer engineer, it would be enough to know one general purpose mathematical package like matlab or mathematica or whatever (I don't know too well this market), one general purpose programming language (I would always recommend C++ over all: FORTRAN, Java or Python), and familiarity with some engineering packages (from drafting and design packages like AutoCAD and CATIA to analysis packages like ANSYS, Fluent, etc.). However, I would not regard the last category (engineering packages) as an absolute must. If you have some smarts and computer knowledge, you can always pick them up very easily---in fact, that, precisely is their selling point anyways!! But learning programming language and learning to use it effectively for numerical analysis is a different story. I think mechanicians should know both these aspects really well in today's world to appreciate technological developments and changes around them, even if they don't eventually do research themselves.

Just my two cents, without thinking too deeply about it, though...

Thanks a lot for your descriptive explanation and the references thet you cited :)

Amit Pandey's picture

The concept of yield point is very loosely defined so far..this manuscript brings a fresh look to this old debatable question.  

Subscribe to Comments for "Concept of Yield Point"

Recent comments

More comments


Subscribe to Syndicate