Definition of Steel Young Modulus 207Gpa 8

[antidio]

Dear all,

Could somebody tell me how was calculated/defined the "standard' value of 207Gpa as Yong Modulus of the steel.
In other words, if I have the stress strain curve as a results of a tensile test (ASTM A370) on a steel sample, then:
1) which part of the curve do I have to consider to calculate the YM. (I mean between which strain limits)
2) which mathematical method do I have to use? (linear regression, chord, etc)

what was historically done to get the value 207 Gpa?

http://www.eng-tips.com/

thanks a lot.

 

[antidio]

Ed, Mag,

I just can confirm that once the load/strain line deviates from straight (YM change) you are probably at 0.1% or less elongation.

this was the reason why we decided to use the SS curve up to 50% of the SMYS to calculate the YM.
In our case 50% of SMYS is about 225Mpa that correspond to about 0.1% strain (assuming YM=210Gpa).


Hey guys I really appreciated all your contribute but one of my initial question remained nor=t replied.

What was historically done to get the value 207 Gpa? I mean how the value of 207Gpa or 200Gpa was calculated. I mean considering which steel and which SS limits?

 

[Maui]

The values of the elastic moduli depend primarily on the types of interatomic bonds that are present and the crystal structure. Several years ago I wrote an FAQ on this very subject. You can read it here:

faq330-1441

Maui

 

[FennLane]

I agree with the comments about modulus depending on interatomic bonding and the way it dominates behaviour.

The definition of yield point has always been the subject of much debate and argument with regard to the correct definition.

With steels that exhibit discontinuous yield it is quite straightforward apart form do you take the Upper or Lower Yield points.

The use of a 0.2% proof stress has been historically used because it was relatively easy to measure and quite repeatable.

It is also a similar level of strain to the discontinuous yield point of a low carbon steel.

Expressions such as limit of proportionality and the like were all developed in the days when strain measurement accuracy was poor and it reality we are potentially arguing about the difference in the thickness of the line produced by an old fashioned X-Y Recorder.

The reality becomes what we really mean by yield point for materials that exhibit a 'continuous' yield and there is no clear 'point'.

Do we mean when one dislocation slips by one Burgers Vector or may be 100 dislocations of may be 1000.

The definition of yield point could be considered to be very closely related to the accuracy of strain measurement and for most conventional materials changing the point we use will only serve to penalise a material and potentially increase cost.

For expedient reasons 0.2% strain was chosen and this has only become an issue with the development of high strength materials that are brittle elastic in nature and mean that 'yields' need to measured at very low strain levels.

If we design using a simple strength of material approach for this type of material we run the risk of experiencing catastrophic failure as the design approach used for materials that are more ductile and can work harden may not be good enough for very strong materials.

As Brittle Elastic materials don't have much work hardening capability this will seriously impact on their ability to tolerate defects and I would imagine that an LEFM approach may be better and evaluating these types of material and KIC may be a better guide to is performance.

I realise this is off topic but yield point measurement is only part of the picture.

 

[MagBen]

What seemingly learned from this thread is that, unlike lots of resources state, steels do NOT exhibit linear-elastic behavior even under a load much lower than .2% yield.

Due to more or less plastic deformation and dislocation movement, modulus seems to decrease with increasing loads (refer to metengr's link). To accurately determine the physics modulus, a regression may be applied, and the load is extrapolated to zero to get the interception as the “true” modulus.

A question arises, who cares about this “true” modulus in engineering world?

 

[FennLane]

Even a regression analysis will be problematic as is there is a measurement error in the basic loading line you will just be fitting incorrect data.

There will always be some small degree of hysteresis in a loading/unloading curve even in a notionally elastic region although this is commonly due to extensometer behaviour.

A traditional extensometer will rock on its knife edges and have some movement and even non-contacting devices can have backlash in their drives.

I haven't used a speckle interferometry laser device so I couldn't comment about this style of extensometer but this may be an improvement on more common devices.

I think accurate determination of E for a typical category of material is important to know and would always use ultrasound or resonance methods of its determination.

To measure E on a day to day basis as part of a QA test just seems a waste of time as it will never be correct to better than a few percentage points.

As was pointed out by Maui there will be little change due to the majority of variables so as long as we have a useable number for design work all should be good.

I was involved in reviewing a programme of JIC Single Specimen tests carried out on A533B specimens some years ago and the crack lengths were calculated using specimen compliance based on an E value of 190GPa which seemed a little on the low side.

Modulus had been measured using an old Hounsfield Tensometer with a Mercury Cell and crosshead deflection and was, IMHO, a complete waste of time.

It took quite some time to re-calculate the results of more than 200 tests with more appropriate values.