Laboratory strength or Field strength?

[Chen1]

Hello There,

I want to model a room and pillar mine, so my task is to ensure that the pillars are safe and stable. I will use Abaqus software,

In ABAQUS, i need to input the material properties, i have the material properties for laboratory samples (2 in high, 1 in diameter). As you know, the strength of the laboratory rock samples is much higher than the in-situ strength.

the question is: what kind of input of parameters should i input in abaqus, is it the properties from the lab. or the laboratory from the field. i will reduce the laboratory strength to get the field strength.

Thanks

 

[oldestguy]

Not that I have designed what you are attempting, I question your sample sizes as being suitable. For some project that involves protecting lives of humans and expensive machinery, the use of what I would term, fourth class samples, seems very unwise, even negligent. The least I would use are "B" size core samples, or larger if possible. Perhaps tell the mine type for better replies from this group.

 

[Mccoy]

Chen1, if by lab streght you mean the strenght of the intact rock and by field strenght you mean the strenght of the rockmass with its joints, then there are the usual methods to get the rockmass strenght. I use very often the roclab freeware from rocscience which I find very convenient to use.

An example in one of my recent projects, a fractured sandstone with lab uniaxal compressive strenght of 35 Mpa turns out to have a rockmass, field strenght of just 1.5 Mpa. This pretty much agrees with oldestguys's observations, the smaller the lab sample is, the less likely it is that it encompasses joints or 'weak spots'.

Of course a specific site survey is necessary besides lab tests. Or extrapolation of rockmass quality by drill cores.

 

[Chen1]

First of all, thanks for your replies.

Well, that means we should input the rockmass properties get more reliable results. not the laboratory test results for the intact cores.

 

[IRstuff]

At a minimum, given the variation in strength, one would think that you would use a gigantic safety factor. I question why the question even came up. Laboratory measurements are done in a controlled environment, which is not reflective of the actual environment.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Chen1]

You know in mining field test on on underground pillar is very rare, so the only option to get an idea about the mechanical properties of rock is to drill core and get some samples, then test these samples. As Mccoy mentioned we might use the rock quality designation of the rock get an idea about quality of the rock mass.

Mining will not be economic if we use a big safety factor. and as you know there is uncertainty about the properties of the rock mass since we depend only on small rock samples.

So i need to scale the laboratory values to get the field values.

Again, these [field values] not the [intact values] should be the input in the numerical modeling, right?

 

[oldestguy]

The idea of trying to model something on the basis of modifying lab results seems to me as only a start. I would think the results of a geologic investigation, using these cores, as well as historical and local information, as to the direction of bedding planes and fracture directions, then possibly modified on the job on the basis of geologic observations as the work moved forward, should narrow down your decisions. In other words, as in tunneling, variations in what is found affects what you do at these differencing situations. And there always is the option of back figuring the calculations based on what you actually see on the job to modify your future predictions.

It still might be of help to know the kind of mine for getting comments from this thread.

 

[IRstuff]

"As you know, the strength of the laboratory rock samples is much higher than the in-situ strength."

Why are you still asking about lab strength? Your caves and pillars are not in the lab, are they?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Chen1]

No the pillars are not in the lab. And we can not test the pillar in the field.

I agree with Oldestguy, that the back analysis might be one of the best approaches to get representative rockmass properties.

 

[ebenozen]

Why are you using Abaqus? There are other methods to evaluating pillar stability. If you want to look at stresses to get an idea of how much the pillars will yield, then this is the right tool. To do this, in addition to the discussed compressive strength and shear strengths, you also need elastic properties (lab and rock mass). I'm guessing you don't have this information. Finite element analysis is typically done to evaluate the impact of abutment stresses and such.

There are other methods to determine if the pillars will be "safe and stable". You can calculate a pillar factor of safety, but you need the following information: load carrying capacity of the pillar (this is the strength of the pillar and is calculated using various methods, not necessarily Roclab), the tributary area that the pillar supports, an estimate of the geometry of the pressure arch and the peripheral load transfer distance (to calculate the load that the pillar will have to carry as mining sheds the load onto peripheral pillars). You may also want to look at abutment stresses or use a high factor of safety to account for an abutment from mining on other levels for example.

 

[Chen1]

As you know there are some empirical method for pillar design, however those methods does not put into account important parameters such as the contact properties between the pillar and both the roof and floor. Moreover it does not account for different geological conditions.

Any i guess Abaqus would be my best bet. the development load can be calculated from the tributary area, and the abutment load can be estimated depending on the abutment angle, however in order to calculate the Safety factor you need to determine the strength, so the rockmass properties are necessary.

ebenzone are you familiar with Abaqus?

Thanks

 

[ebenozen]

I don't know the program Abaqus, but I am familiar with finite element analysis. The only FEA software I'm familiar with is FLAC and by no means am I an expert with it.

I think you're better off staying away from finite element analysis for your project. As you mentioned, there are other input parameters you need to come up with to run Abaqus and uncertainty in these parameters can lead to nebulous results. I think it's best to simplify the analysis and use input parameters that you can determine with greater confidence.

Best of luck.