Artificial Neural Networks(ANN)- Basics


Artificial Neural Networks are a powerful tool for the purpose of prediction and recognition of patterns. This just a introduction to Neural Networks in a compacted manner. It is advised to have a more detailed study after going through the following article.

For starters, you need a good set of programming skills in C or C++ to design a neural network. And if you are designing using MATLAB, no such programming skills are required, Neural Networks tool is pre-embedded in it.

When Neural Networks are Used

The basic purpose of a neural network is to predict certain output data for a corresponding input data. For example consider a case in manufacturing industry where the productivity depends on the efficiency of worker, efficiency of machine, number of workers working and other miscellaneous factors. We have the data for the past 24 months and need to predict the productivity for the next month, in such a situation we use the help of Neural Networks to predict its value. Here Productivity will be the desired output while others factors mentioned above would be the inputs. So if we give the value of these inputs to a custom designed neural network, we would obtain a output.

BASICS

A neural network has three basic layers: an input layer, a hidden layer (can be more than one) and an output layer.

download

 

Data is processed in a sequential order in different layers and desired output is obtained only after a proper network is designed.

* Each circle in the above picture is called a node or a neuron and each layer has many number of nodes depending on our choice. All computations are done inside the nodes and processed data is carried forward for the next set of computations.

* The computation in each of the nodes is done using a function called Activation function. There are a number of activation functions, for example sigmoid function, and the choice depends on the work.

* In the above figure, we can see that each node in a layer is connected several other nodes in the previous layer. The nodes in the previous layer are connected to present ones using weights and biases. The following picture explains it.

images

 

Here f(u) is the activation function.

*Another important parameter is the selection of training algorithm, which is used in the training process of a network and which will be discussed later. There a number of training algorithms and the most popular one is a Back Propagation Algorithm. Each one has its own pros and cons.

* The selection of type of network also has an impact on the efficiency of the network. One example is of a type of network is a feed-forward network. The type of network chosen comes into picture when a network is run after training.

PROCEDURE FOR DESIGNING A NEURAL NETWORK

There are three basic steps involved: Training, Validation and Prediction. The available data should be divided accordingly. A thumb rule would be to have 80% of available data for training and remaining 20% for validation. The steps involved in designing a neural network are:
1. Fixing the structure of the network: choosing the number of hidden layers, number of nodes in each layer and activation function for each layer.

2.  Giving initial weights and biases if designing through the use of programming languages.

3. Giving the required data and training the network.

4. Once training of a network is done, its validation must be done with the remaining set of data. This is an important step since the efficiency of network can be determined based on this step. If desirable results are not obtained the network must be re-designed and trained again and re-validated. So it is a trail and error method.

5. Once validated and the possible range of errors is established, the network can be run to predict the value and it should be represented with the possible percentage of error which in normal cases lied around 10%.

NOTE: Increasing the number of nodes in a layer and the number of layers allows the network to solve more complicated problems, but it also increases the computational space required, so a balance must be maintained.

Advertisements

Composition and Hydration reactions of Fly ash used in Concrete


A. Introduction

Fly ash is one of the most extensively used by-product materials in the construction field resembling Portland cement. It is an inorganic incombustible, finely divided residue collected or precipitated from the exhaust gases of any industrial furnace.

Fly ash is extracted from flue gases through Electrostatic Precipitator in dry form. It is a fine material & possesses good pozzolanic property.

Fly ash produced in modern power stations is of good quality as it contains low sulphur & very low unburnt carbon i.e. less loss on ignition. In order to make fly ash available for various applications, most of the new thermal power stations have set up dry-fly ash evacuation & storage system.

B. Chemical Composition

The composition varies with type of fuel burnt, load on the boiler and type of separation. The fly ash consists of spherical glassy particles ranging from 10 to 15 micron in diameter. The constituents of fly ash are mentioned below.

Silicon dioxide               —– SiO2               — 30 – 60 %

Aluminum oxide           —– Al2O3           — 15 -30 %

Unburnt fuel                    —– (Carbon)     — up to 30 %

Calcium oxide                 —– CaO                — 1-7%

Magnesium oxide          —– (MgO)           — small amounts

Sulphur trioxide             —– (SO3)            — small amounts

C. Hydration Reactions when Fly ash is used in OPC

As mentioned in section 2; OPC is made up of four principal mineralogical phases symbolically represented by C3S, C2S, C3A and C4AF. The hydration reactions of these chemical compounds as mentioned in the respective section are as follows:

For C3S:

2C3S  +  6H        —> C3S2H3 + 3Ca(OH)2

For C2S:                                                                                            

2C2S  +  4H        —>   C3S2H3 + Ca(OH)2

For C3A:

C3A   +  6H        —>   C3AH6

One of the primary benefits of fly ash is its reaction with available lime and alkali in concrete, producing additional cementitious compounds. The following equations illustrate the pozzolanic reaction of fly ash with lime to produce additional calcium silicate hydrate (C-S-H) binder:

Cement Reaction       :           C3S/ C2S   +    H   →   C-S-H + Ca(OH)2

Pozzolanic Reaction  :            Ca(OH)2   +    S    →   C-S-H

S —- Silica from Fly ash constituents

So, clearly from the above equation we can interpret that the excess lime content produced by the hydration reactions of cement, which weakens cement by leaching and other processes, is reduced by the use of silica. Also due to the formation of additional calcium silicate hydrate (C-S-H) binder, the strength of the mix as a whole increases.

D. Advantages of using Fly ash

The benefits of using Fly ash along with OPC in mix design are as follows:

  • Improved workability
  • Decreased water demand
  • Reduced heat of hydration
  • Increased ultimate strength
  • Reduced permeability
  • Improved durability

 

Micro structure of Concrete


Research has made it possible to significantly improve the strength and durability of concrete without a great increase in cost. As a result of the research “Concrete technology” has under gone from a macro to a micro level study in the enhancement of strength and durability. By bringing about changes in the micro structure of concrete, strength of concrete as a whole can be improved.

Magnetic resonance imaging (MRI) of a concrete bloc

Magnetic resonance imaging (MRI) of a concrete bloc

The Magnetic Resonance Imaging (MRI) of a concrete block clearly indicates three components of which concrete is made of. Hence the three components of which a concrete hardened block is made up are:

1. Hardened Cement Paste (HCP)

2. Aggregate

3. Transition zone: Weakest zone (Extensive research is being made to improve properties of this zone)

Summary:

Knowledge of the micro structure and properties of the individual components of concrete and their relationship to each other is useful for exercising control on the properties of Concrete.
• It is possible to modify the properties of concrete by making suitable changes in the micro structure.

 

pH of water (Environmental Engineering)


pH scale

pH scale

pH of water sample is an important parameter which is to be constantly checked for or monitored either for performing the various tests on water samples or for the release of waste water into streams or rivers. pH plays an important role as this value would affect the solubility of certain nutrients and metal compounds.

* pH is defined as the negative of logarithmic of hydrogen ion concentration of a sample. Mathematically represented as

pH =  – log [ H+]                     where H+ represents the hydrogen ion concentration

* It defines how acidic or basic the sample is. It varies from 0-14 where 0 represents the sample to be very highly acidic and 14 being a very alkaline one.

* In case of water either for testing or for the release into streams or rivers a pH value in the range 6.5-8.1 is to be maintained as only in this range of pH aquatic life would sustain and also the wide range of indicators etc we will be using in the tests are calibrated for this range of pH. If the pH value decreases due to several reasons of a river or a stream like acid rain etc this will certainly increase the solubility of heavy metals which in turn increases the toxicity of water. Also a change in pH value varies the concentration of various nutrients in the water which will affect the growth of aquatic plants changing the concentration of “Dissolved Oxygen“. 

* Various tests can be performed in the laboratory to determine the pH value of a water sample. Some of them are:

1. pH determination using pH paper

2. pH determination by color comparison (Universal Indicator is used)

3. pH determination by using pH meter.

Why Bio Chemical Oxygen Demand (BOD) at 5 days??


BOD or “Bio Chemical Oxygen Demand” is an important factor which describes us how pure the given water sample is. It represents the amount of Dissolved Oxygen consumed by microbes to convert the organics present to carbon dioxide and water. Higher the BOD value higher is the water polluted. A sample of water with lesser BOD value is less polluted than a one with much higher value.

* Generally as per global standards BOD value of sample at 5 days which incubated at 20 degrees Celsius is represented as the standard BOD of the sample. The reason for taking 5 days BOD as standard is due to the following fact:

The “Dissolved Oxygen” present in water gets used for the decomposition of organic matter as well as nitrogenous matter.

From the combined BOD curve, we have two oxygen demands: 1. Carbonaceous BOD (initial) , 2. Nitrogenous BOD (final).

The first 5 days demand is the “Carbonaceous Bio Chemical Oxygen Demand” which accounts for 68% of the total BOD value. Hence taking the value of 5 days BOD as a standard one represents the oxygen demand for the decomposition of organic matter only.

 

What is meant by “Characteristic strength” (fck) of concrete??


Characteristic strength of concrete is one of the important properties of concrete which indeed unanimously by design engineer or any other person involved in the construction sector.

The compressive strength of concrete is given in terms of the characteristic compressive strength of 150 mm size cubes tested at 28 days (fck)- as per Indian Standards (ACI standards use cylinder of diameter 150 mm and height 300 mm). The characteristic strength is defined as the strength of the concrete below which not more than 5% of the test results are expected to fall.

This concept assumes a normal distribution of the strengths of the samples of concrete.

Normal Distribution curve on test specimens for determining compressive strength

Normal Distribution curve on test specimens for determining compressive strength

The above sketch shows an idealized distribution of the values of compressive strength for a certain number of test specimens. The horizontal axis represents the values of compressive strength in MPa. The vertical axis represents the number of test samples for a
particular compressive strength. This is also termed as frequency.

The average of the values of compressive strength (mean strength) from the graph is 40 MPa. The characteristic strength (fck) is the value in the x-axis below which 5% of the total area under the curve falls. From the graph we can clearly say that 30 MPa is the characteristic strength of the given concrete mix. The value of fck is lower than fcm (40 MPa- mean strength) by 1.64σ, where σ is the standard deviation of the normal distribution.

So we can say the given concrete mix has a characteristic strength of 30 MPa or it is a M30 grade mix.

   M- Mix

* Note: For a 95% confidence level k=1.64 , hence k value varies on the confidence level of the experiment

Definition:

Characteristic strength of concrete is the strength of concrete specimens casted and tested as per given code of practice and cured for a period of 28 days; 95% of tested cubes should not have a value less than this value.

Jar test for determining “Coagulant” dosage in Water treatment


Image

Jar test apparatus

Aim:

To determine the optimum dosage of coagulant to remove small or charged particles present inside water by using “Alum” as coagulant.

Principle :

The two basic terms which can exactly explain the happenings of this experiment are “Coagulation” and Flocculation“.

1. Coagulation: It is the process of addition of a chemical to de-stabilize a stabilized charged particle.

2. Flocculation: It is a slow mixing technique which promotes agglomeration and helps the particles to settle down.

* Generally we encounter very fine and  charged clay like particles in water treatment which should be removed before we continue for further processes. These impurities do not settle by gravity when the water is passed through a sedimentation tank. The reason being that these are charged particles, they repel each other and just stay.

* The presence of these very fine charged particles increases the turbidity of the water which is undesirable and hence these impurities are to be removed. Therefore which will be using a chemical which dissociates as soon as it added to water and helps in the process of “Coagulation”. In the present experiment we are using “Alum”  [Al2(SO4)3. 18H2O] as the clarifying agent.

*When alum solution is added to water, the molecules dissociate to yield  SO4^2- and Al3+. These charged species combine with the charged colloidal particles to neutralize the charge. A detailed explanation of the charge removal can be found in the web which will be based on two basic definitions Stern potential and Zeta Potential.

* Through the slow mixing or so called “Flocculation” a process known as agglomeration occurs which combines the charged particles into a compact whole and helps in the settling of the particle. That is the reason why we have step of “slow mixing” in the present experiment.

Apparatus:

1. Jar testing apparatus

2. Turbidity meter

3. Beaker, burette, pipette

Reagents required: Alum solution (1 ml containing 10 mg of alum)

Procedure :

1. Take 1000 ml of given sample in 6 beakers.

2.  Find the pH of the sample and adjust it to 6 to 8.5.

3. Now add 1 ml, 2 ml, 4 ml, 8 ml, 10 ml, 12 ml of alum respectively in each one of the beakers.

4. Now insert the paddle of the jar testing apparatus inside the beakers and start it.

5. Initially maintain a speed such that the paddles rotate at an angular velocity of 100 rpm for a time of 1 minute.

6. Now adjust the speed such that the paddles rotate at 40 rpm/min for a time of 9 minutes.

7. Now allow the beakers to settle down for 10 minutes.

8. Make an observation as of which of the 6 beakers is most clearer. Also measure the turbidity of each beaker using a turbidity meter and tabulate your results.

9. Plot a graph “Settled turbidity” Vs “Coagulant dosage”.

Result: 

The optimum value of coagulant dosage from the graph should be reported.

* The optimum value of coagulant generally lies between 8 to 10 for normal water from rivers.

The graph varies as shown (only shape not exact graph)

Graph