1-THEORY 1.

1-PROCESS CONTROL Process control refers to the methods that are used to control process variables when manufacturing a product. For example, factors such as the proportion of one ingredient to another, the temperature of the materials, how well the ingredients are mixed, and the pressure under which the materials are held can significantly impact the quality of an end product. Manufacturers control the production process for three reasons(1): Reduce variability Increase efficiency Ensure safety1 In controlling a process there exist two type of classes of variables(2). 1. Input Variable This variable shows the effect of the surroundings on the process. It normally refers to those factors that influence the process. An example of this would be the flow rate of the steam through a heat exchanger that would change the amount of energy put into the process. There are effects of the surrounding that are controllable and some that are not. These are broken down into two types of inputs. a. Manipulated inputs: variable in the surroundings can be control by an operator or the control system in place. b. Disturbances: inputs that can not be controlled by an operator or control system. There exist both measurable and immeasurable disturbances.

2. Output variable- Also known as the control variable These are the variables that are process outputs that effect the surroundings. An example of this would be the amount of CO2 gas that comes out of a combustion reaction. These variables may or may not be measured. As we consider a controls problem. We are able to look at two major control structures. 1. Single input-Single Output (SISO)- for one control(output) varible there exist one manipulate (input) variable that is used to affect the process
1

2. Multiple input-multiple output(MIMO)- There are several control (output) variable that are affected by several manipulated (input) variables used in a given process(2).

1.1.1- Transfer Functions A Transfer Function is the ratio of the output of a system to the input of a system, in the Laplace domain considering its initial conditions and equilibrium point to be zero. If we have an input function of X(s), and an output function Y(s), we define the transfer function H(s) to be(3):

(1)

Figure 1.1 :Block diagram of Transfer functions For comparison, we will consider the time-domain equivalent to the above input/output relationship. In the time domain, we generally denote the input to a system as x(t), and the output of the system as y(t). The relationship between the input and the output is denoted as the impulse response, h(t). We define the impulse response as being the relationship between the system output to its input. We can use the following equation to define the impulse response:

(2) Impulse Function It would be handy at this point to define precisely what an "impulse" is. The Impulse Function, denoted with (t) is a special function defined piece-wise as follows:

(3) The impulse function is also known as the delta function because it's denoted with the Greek lower-case letter . The delta function is typically graphed as an arrow towards infinity, as shown below:

Figure 1.2 : mpulse (delta) function 1.1.2- Step Response Similarly to the impulse response, the step response of a system is the output of the system when a unit step function is used as the input. The step response is a common analysis tool used to determine certain metrics about a system. Typically, when a new system is designed, the step response of the system is the first characteristic of the system to be analyzed.However, the impulse response cannot be used to find the system output from the system input in the same manner as the transfer function(3).

1-2 DYNAMIC BEHAVIOURS OF FIRST ORDER AND SECOND ORDER SYSTEMS 1.2.1 First-Order Systems A one-degree-of-freedom first-order system is governed by the first-order ordinary differential equation(4,5,6)

(4) where y(t) is the response of the system (the output) to some forcing function F(t) (the input). Eq. (4) may be rewritten as

(5) where =a1/a0 has the dimension of time and is the time constant for the system and k =1/a0 is the gain.

Response of a First-order System to a Step Input Consider a first-order system subjected to a constant force applied instantaneously at the initial time t = 0 (4,5,6)

(6) The initial condition is y(0) = 0. The solution to Eq. (5) with the step input Eq. (6) is then

(7) The response approaches the final value y= kA exponentially. By using the boundary conditions equation (7) then may be rewritten as

(8) The rate at which the response approaches the final value is determined by the time constant. When t = , y has reached 63.2% of its final value as illustrated in Figure 3. When t =5, y has reached 99.3% of its final value.

Figure 1.3 : First Order systems The time constant of a system can be determined from the measured response using a linear regression. Taking the natural log Eq. (8) yields

(9) The slope s of the natural log term plotted against t gives the time constant through the relation s = -1/(4,5,6).

Transient Response of a Thermocouple The dynamic response of a sensor is often an important consideration in designing a measurement system. The response of a temperature sensor known as a thermocouple (TC) may be modeled as a first-order system. When the TC is subjected to a rapid temperature change, it will take some time to respond. If the response time is slow in comparison with the rate of change of the temperature that you are measuring, then the TC will not be able to faithfully represent the dynamic response to the temperature fluctuations(6). A model of the response of a TC is based on a simple heat transfer analysis. The rate at which the sensor exchanges heat with its environment must equal the rate of change of the internal energy of the sensor. If the dominant mechanism of heat exchange is convection (neglecting conduction and radiation), as it is for a TC in a fluid, then this energy balance is

(10) h is the convection coefficient, A is the surface area of the sensor, T is the temperature, m is the TC mass, and c is the heat capacity. Writing Eq. (11) in the form of Eq. (5)

(11) where the time constant is

(12) 1.2.2 Analysis Of Second-Order Systems A second-order system is one whose output, y(t), is described by a second-order differential equation. For example, the following equation describes a second-order linear system(7):

(13) If ao 0, then Equation (13) yields

(14) Equation (14) is in the standard form of a second-order system, where = natural period of oscillation of the system = damping factor K = steady state gain The very large majority of the second- or higher-order systems encountered in a chemical plant come from multicapacity processes, i.e. processes that consist of two or more first-order systems in series, or the effect of process control systems. Laplace transformation of Equation (14) yields

(15) Case A: (over-damped response), when > 1, we have two distinct and real poles. In this case the inversion of Equation (15) by partial fraction expansion yields

(16) Where cosh(.) and sinh(.) are the hyperbolic trigonometric functions defined by

(17) Case B: (critically damped response), when = 1, we have two equal poles (multiple pole). In this case, the inversion of Equation (15) gives the result

(18) Case C: (Under-damped response), when < 1, we have two complex conjugate poles. The inversion of Equation (15) in this case yields
7

(19)

Figure 4 : Underdamped Systems - Overshoot: Is the ratio of a/b, where b is the ultimate value of the response and a is the maximum amount by which the response exceeds its steady state value. It can be shown that it is given by the following expression:

(20) - Decay ratio: Is the ratio of the amount above the stead state value of two successive peaks, c/a. it can be shown that it can be calculated by the following equation:

(21)
8

- Rise time: tr is the the process output takes to first reach the new steady state value. - Time to first peak: tp is the time required for the output to reach its first maximum value. - Settling time: ts is defined as the time required for the process output to reach and remain inside a band whose width is equal to 5 % of the total change in the output. - Period: Equation (21) defines the radian frequency, to find the period of oscillation P (i.e. the time elapsed between two successive peaks), use the well-known relationship = 2/P;

(22)

2. EXPERIMANTAL METHOD The experimental set-up consists of different U-manometers in different diameters and that contains diffrent type of liquids via their properties such as water, glycerol and their mixtures. The pressure difference in the U-manometer was created by a vacuum generator. 2.1.DESCRIPTION OF APPARATUS

Figure 2.1. U tube manometer[8] 2.2. EXPERIMENTAL PROCEDURE Pressure difference was applied on the U-manometer by vacuum generator and determine the variation of the liquid level with time until the manometer balanced. The vacuum pump was stoped when the constant liquid level was observed. This process was repeated for all overdamp U-manometer, and determine again the variation of the liquid level with time. For underdamped U manometer the vacuum generator was opened and then oscilation was observed . The liquid level and their times was determined for step and impulse function.
10

3.0 RESULTS AND DISCUSSION 3.1 U TUBE MANOMETERS Table 3.1 U tube manometers properties properties (g/cm3) (Cp) D (cm) L( cm) (s) Manometer Manometer 1 0,885 137,6 0.6 88 0.212 14,64 2 0,997 0,894 1,1 95 0.220 0,026 Manometer 3 1,261 902,85 0.6 102 0.228 72,52 0,885 137,6 1,10 98 0.224 4,6 1,058 1,362 1,10 85 0.208 0,0354 1,261 902,85 1,10 116 0.243 23,03 Manometer 4 Manometer5 Manometer 6

According to Table 3.1 the viscosity of liquid in manometer 2 and 5 were realy smaller than other and their diameter were same or bigger. This conditions effected the damping factor to be smaller than 1. As a result manometer 2 and 5 could not absorb the effect of disturbition like others so that their response will to be underdamped conditions. To determine the response time we must look their time constant. The time constant was proportional with square root of their lenght.As a result manometer 4 and 6 had a fast response time. 3.2. RESULTS for OVERDAMPED U-MANOMETERS Table 3.2.1 Experimental Responses of Overdamped U-manometers to step change
Manometer 1 t(S) 0 3 6 9 12 15 18 21 24 27 t/ 0,000 14,151 28,302 42,453 56,604 70,755 84,906 99,057 113,208 hr/Kp 0,000 0,415 0,701 0,844 0,918 0,952 0,980 0,993 1,000 hf/Kp 1,000 0,510 0,238 0,143 0,068 0,041 0,007 0,000 Manometer 3 t/ 0,000 13,158 26,316 39,474 52,632 65,789 78,947 92,105 105,263 118,421 hr/Kp 0,000 0,430 0,645 0,766 0,850 0,916 0,944 0,972 0,991 1,000 hf/Kp 1,000 0,589 0,336 0,206 0,131 0,075 0,047 0,019 0,000 Manometer 4 t/ 0,000 13,393 26,786 40,179 53,571 66,964 80,357 93,750 hr/Kp 0,000 0,420 0,623 0,754 0,841 0,884 0,928 1,000 hf/Kp 1,000 0,594 0,319 0,145 0,072 0,029 0,014 0,000 Manometer 6 t/ 0,000 12,346 24,691 37,037 49,383 61,728 74,074 hr/Kp 0,000 0,571 0,771 0,886 0,943 0,971 1,000 hf/Kp 1,000 0,457 0,229 0,114 0,057 0,029 0,000

11

According to table 3.2.1 as predicted at table 3.1.1 fast response occured in manometer 4 and 6. Because of the tube lenght and diameter of the tube was bigger than other tubes so that manometer 6 can easily absorp the effect of distirubition and give us fast response. But manometer 4 must had a fast response time because its viscosity was smaller than manometer 6s liquid maybe some personal mistake in the experiment.

1,200

1,000

0,800 M1 M3 M4 0,400 M6

hr/Kp

0,600

0,200

0,000 0,000

20,000

40,000

60,000 t/to

80,000

100,000

120,000

140,000

Figure 3.2.1. Experimental hr/kp versus t/ values According to Figure 3.2.1 we can determine the response time . Kp values were the ultimate values. Manometer 6 was reach their ultimate values faster than others when fluid was rising.

12

1,200 1,000 0,800 hf/kp 0,600 0,400 0,200 0,000 0,000 M1 M3 M4 M6

20,000

40,000

60,000 t/to

80,000

100,000 120,000

Figure 3.2.2. Experimental hf/Kp versus t/ values When the fluid was falling again the manometer 6 had a fast response time others .

Table 3.2.2 Theoretical responses of Overdampded U-manometers to step change

Manometer 1 t(s) 0 3 6 9 12 15 18 21 24 27 t/ 0,000 14,151 28,302 42,453 56,604 70,755 84,906 99,057 113,208 hr/Kp 0,000 0,384 0,620 0,811 0,891 0,938 0,964 0,979 0,988 hf/Kp 1,000 0,616 0,380 0,189 0,109 0,062 0,036 0,021 0,012 Manometer 3 t/ 0,000 13,158 26,316 39,474 52,632 65,789 78,947 92,105 105,263 118,421 hr/Kp 0,000 0,087 0,166 0,238 0,304 0,365 0,420 0,470 0,516 0,558 hf/Kp 1,000 0,913 0,834 0,762 0,696 0,635 0,580 0,530 0,484 0,442 Manometer 4 t/ 0,000 13,393 26,786 40,179 53,571 66,964 80,357 93,750 hr/Kp 0,000 0,771 0,948 0,988 0,997 0,999 1,000 1,000 hf/Kp 1,0000 0,2290 0,0524 0,0120 0,0027 0,0006 0,0001 0,0000 Manometer 6 t/ 0,000 12,346 24,691 37,037 49,383 61,728 74,074 hr/Kp 0,000 0,238 0,419 0,558 0,663 0,743 0,804 hf/Kp 1,000 0,762 0,581 0,442 0,337 0,257 0,196

This table show the theoretical responses of overdamped u manometers t/ values must be same with the experiment . hr/Kp values were different with experimental because of the persanol mistakes.

13

1,2 1 0,8 hr/kp 0,6 0,4 0,2 0 0 50 t/to 100 150 M1 M3 M4 M6

Figure 3.2.3. Theoretical hr/Kp versus t/ values M1 and M4 included same fluid and their viscoty values were smaller so that their response times must be faster than others and also M6s lenght was bigger than M3 so that M6 must gives us fast response time.
1,2 1 0,8 hf/Kp 0,6 0,4 0,2 0 0 50 t/to 100 150 M1 M3 M4 M6

Figure 3.2.4. Theoretical hf/Kp versus t/ values Same approach with the Figure 3.2.3 when the fluid was falling
14

3.3 RESULTS FOR UNDERDAMPED U-MANOMETERS (TO STEP CHANGE) Table 3.3.1 Period of Oscillation and Radian Frequency of Underdamped U-Manometers Manometer Manometer 2 Period of Oscillation T(s) Radian Frequency W(s) 1,383 4,544 5 1,33 4,802

Period of oscilation of manometer 2 and 5 were nearlly close together but manometer 2 little bit long. The reason maybe the viscoty of liquid in manometer 2 was small so it rised more than manometer 5 and that effected the raidan frequency .

Table 3.3.2 Experimental Responses of Underdamped U-manometers to Step Change Manometer 2 texp(s) 0 1 2 3 4 5 6 7 8 0,000 1,510 1,780 2,590 3,230 3,960 4,900 5,810 6,890 t/ 0,000 6,864 8,091 11,773 14,682 18,000 22,273 26,409 31,318 h/Kp 0,000 1,000 0,522 0,882 0,676 0,809 0,728 0,699 0,743 Manometer 5 texp(s) 0,000 1,430 1,890 2,470 3,460 4,220 5,190 6,110 7,160 t/ 0,000 6,875 9,087 11,875 16,635 20,288 24,952 29,375 34,423 h/Kp 0,000 1,000 0,500 0,890 0,646 0,768 0,720 0,720 0,744

As we expected the h /Kp values shows us the oscillation was occured because of the their damping factor . And also the input was step function so that the osicalliton was reach one point
15

Table 3.3.3 Theoretical Responses of Underdamped U-manometers to Step Change Manometer 2 ttheo(s) 0 1 2 3 4 5 6 7 8 9 10 0,327 1,018 1,708 2,398 3,088 3,778 4,468 5,158 5,848 6,538 7,228 t/ 1,486 4,627 7,764 10,900 14,036 17,173 20,309 23,445 26,582 29,718 32,855 h/Kp 0,138 1,795 0,265 1,680 0,372 1,581 0,463 1,496 0,541 1,424 0,608 Manometer 5 ttheo(s) t/ 0,417 1,081 1,745 2,409 3,073 3,737 4,401 5,065 5,729 6,393 2,005 5,197 8,389 11,582 14,774 17,966 21,159 24,351 27,543 30,736 h/Kp 0,154 1,740 0,354 1,562 0,513 1,421 0,637 1,311 0,734 1,226

The cause of reading mistakes the experimental values was not close with the experimental values.
2,000 1,800 1,600 1,400 1,200 h/kp 1,000 0,800 0,600 0,400 0,200 0,000 0,000 5,000 10,000 15,000 20,000 25,000 30,000 35,000 t/to M2-exp M2-theo

Figure 3.3.1. Comparison of Experimental and Theoretical Responses for M-2 Experimental values were not correctly readed .
16

2,000 1,800 1,600 1,400 1,200 h/kp 1,000 0,800 0,600 0,400 0,200 0,000 0,000 10,000 20,000 t/to 30,000 40,000 M5-exp M5-theo

Figure 3.3.2. Comparison of Experimental and Theoretical Responses for M-5 Experimental values were not correctly readed .

Table 3.3.4 Comparison of Theoretical and Experiment Overshoot, Decay Ratio and Response Time to Step Change Monometer 2 Monometer 5 Experimental Theoretical Experimental Theoretical Overshoot Decay ratio Response Time 0,187 0,475 6,89 0,922 0,85 7,228 0,196 0,164 7,16 0,897 0,805 6,393

In the experiment the lenght and time was not readed correctly but that should not effect the response time our experimantal values were close with theoretical.

17

3.4. RESULTS FOR UNDERDAMPED U-MANOMETERS (TO IMPULSE CHANGE) Table 3.4.1 Experimental Responses of Underdamped U-manometers to Impulse Change
texp(s) 0 1 2 3 4 5 6 7 8 9 10 0,000 0,650 1,280 1,890 2,550 3,270 4,170 5,010 5,770 6,670 7,870 8,690 t/ 0,000 2,955 5,818 8,591 11,591 14,864 18,955 22,773 26,227 30,318 35,773 39,500 1,000 -0,693 0,511 -0,341 0,295 -0,239 0,295 -0,239 0,193 -0,114 0,091 h/Kp texp(s) 0 1,1 1,79 2,5 3,12 4 4,89 5,75 6,59 7,71 t/ 0,000 5,288 8,606 12,019 15,000 19,231 23,510 27,644 31,683 37,067 h/Kp 0,000 1,000 -0,791 0,674 -0,372 0,186 -0,140 0,070 -0,023 0,012

The input was the impulse function so that the h/Kp values changes positive to negative. The lenight of oscicallation should reach 0.

Table 3.4.2 Theoretical Responses of Underdamped U-manometers to Impulse Change ttheo(s) 0 1 2 3 4 5 6 7 8 9 10 11 -0,346 0,346 1,038 1,730 2,422 3,114 3,806 4,498 5,190 5,882 6,574 7,266 t/ -1,573 1,573 4,718 7,864 11,009 14,155 17,300 20,445 23,591 26,736 29,882 33,027 h/Kp -1,042 0,960 -0,885 0,815 -0,751 0,692 -0,638 0,588 -0,541 0,499 -0,460 0,423 theo(s) -0,333 0,333 0,998 1,663 2,328 2,993 3,658 4,323 4,988 5,653 6,318 6,983 t/ -1,599 1,599 4,796 7,993 11,190 14,387 17,584 20,781 23,978 27,175 30,373 33,570 h/Kp -1,057 0,947 -0,845 0,753 -0,668 0,591 -0,522 0,459 -0,403 0,352 -0,306 0,266

18

12 13 14 15 16 17

7,958 8,650 9,342 10,034 10,726 11,418

36,173 39,318 42,464 45,609 48,755 51,900

-0,390 0,359 -0,331 0,305 -0,281 0,259

1,5

0,5 h/kp

0 0,000 -0,5

M2 exp 10,000 20,000 30,000 40,000 50,000 60,000 M2-theo

-1

-1,5

t/to

Figure 3.4.1. Theoretical and experimental values for M-2 According to Figure 3.4.1 the experimental and theoretical curve was close early but than some of the mistakes maybe reading mistakes was effectted the phase of the oscillation. But both of them was aproach to zero because of the impulse function.

19

1,500

1,000

0,500 h/kp M5-exp 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 M5-theo

0,000 0,000 -0,500

-1,000

-1,500

t/to

Figure 3.4.2. Theoretical and experimental values for M-5 According to Figure 3.4.2 the experimental and theoretical curve was close early but than some of the mistakes maybe reading mistakes was effectted the phase of the oscillation. But both of them was aproach to zero because of the impulse function.

Table 3.4.3 Comparison of Theoretical and Experimental Overshoot, Decay Ratio and Response time to Impulse change Monometer 2 Monometer 5

Experimental Theoretical Experimental Theoretical Overshoot Decay ratio Response Time 0,511 0,577 8,6 0,922 0,85 11,42 0,674 0,276 7,71 0,897 0,805 6,98

20

4. CONCLUSIONS In this experiment ,to determine the effects of liquid properties and shape of U-tube manometers on response time by using step and impulse input, U-manometer systems, which are manometer-1 with engine oil, manometer-2 with water, manometer-3 with glycerol, manometer-4 with engine oil, manometer-5 with 15% glycerol solution and manometer-6 with glycerol were used. In the overdamped systems (m-1,m-3,m-4,m-6), the damping factor was calculated and it was observed that their damping factors were greater than 1. These systems can easily absorb the energy of disturbiton and the reason of this viscosity of liquids that contained in these manometers were high enough according to their diameter and length.Furtheremore, to compare their response time, it was observed that higher length and higher diameter cause the response time to get low for same liquid. In the underdamped systems (m-2 ,m-5), the damping factor was calculated again and it was observed that their damping factor were smaller than 1. As we expected they relased their energy with doing oscillation step by step. Our experimental values was different from the theoretical values.The reason of this the oscillation was realy fast so that the reading mistakes was done. Howewer, according to theoretical and experimental response time, we could observed that the impulse system had a higher response time than step system. The reason of this, while they relasing their energy which comes from disturbition from vacuum generator, the potential energy differences at step function was small than impulse function.

21

5. NOMENCULATURE A, B Constants in the transfer function At Surface area of bulb for heat transfer (m2) g Acceleration of gravity (m/s2) Kp Static gain or gain (m) L Total length of the liquid in U-manometer (m) m Mass of liquid in the monometer (kg) r Liquid lever difference at any time in U-manometer (m) t Time (s) tr Rise time (s) T period of oscillation (s/cycle) Q Volumetric flow rate of the liquid (m3/s) p Time constant (s) Viscosity of the liquid (Pa.s) Density of the liquid (kg/m3) Radian frequency (radian/s) Damping factor

22

6.REFERENCES 1- http://www.pacontrol.com/download/Process%20Control%20Fundamentals.pdf 2https://controls.engin.umich.edu/wiki/index.php/Process_Control_Definitions_and_Terminolo gy 3-. http://en.wikibooks.org/wiki/Control_Systems/Transfer_Functions 4- J.P. Holman, Experimental Methods for Engineers, 7th Ed., McGraw-hill, New York, 2001: First-order systems, p. 19-23; Thermocouples p. 368-377; Linear regression p. 9194; Signal conditioning (RC Circuits) p. 183-190. 5-R.S. Figliola and D.E. Beasley, Theory and Design for Mechanical Measurements, Wiley, New York, 1991, p. 63, 73. 6- Omega Technologies Handbook, Thermocouple Reference Tables, Omega Engineering Inc., 1993, p. B172. 7http://faculty.ksu.edu.sa/alhajali/Publications/Dynamic%20Behavior%20of%20First_Second %20Order%20Systems.pdf 8.http://www.edibon.com/products/?area=fluidmechanicsaerodynamics&subarea=fluidmecha nicsgeneral

23

7. APPENDIX

24