Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

Process dynamics > Definitions > Flashcards

Definitions Flashcards

(64 cards)

Start Studying
1
Q

Benefits of process control

A
  • Lower product variability
  • Achieve set points
  • Energy conservation by preventing waste
  • Reduce emissions
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Why is process control necessary

A

Minimise effect of disturbances

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Sensor

A

Part of primary element that senses a change in the control variable

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Transducer

A

Converts signal from sensor into an electrical signal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Manometer

A

U-Shaped tube filled with liquid of known density. Each side connected to different pressure source. Higher pressure source pushes liquid down and causes it to rise on the other side. Difference in height between two sides gives pressure difference according to ρgh

  • Cheap, rugged, reliable
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Bourdon tube

A

Fluid is pumped into a C-shaped tube. Pressure tries to straighten the curve which causes the gauge’s needle to rotate and display the pressure reading.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Pneumatic DP Cell

A

Consists of two sides, each connected to different pressure source. A diaphragm is then pushed by the higher of the two sources, causing the diaphragm to bend a force bar. The force bar tilts a baffle which control an air signal that is sent to a pressure gauge

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Electronic DP Cell

A

Difference in pressure on two sides of diaphragm. Pressure is exerted through silicone oil which causes a change in the distance between the two plate and so a change in capacitance. The change in capacitance is measured and used to calculate pressure.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Ultrasonic measurement

A

Transducer sends ultrasonic pulse to surface of liquid to be measured. Surface reflects the pulse and the time taken for pulse to return to transducer is measured. U = L/T used to calculate distance

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Bimetal thermostat

A

Made of two different metals that expand at different rates. Difference in expansions causes the strip to bend. At a certain temperature, the faster expanding metal touches a contact which completes an electric circuit and turns on a device, indicating the temperature has been reached.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Thermocouple

A

Made of 2 different metal wires joined at one end (hot end) with the other ends connected to a voltmeter. The hot end is heated and the metals react differently as one gets more excited than the other. Creates a potential difference between the metals with a higher temp. difference between hot and cold ends leading to a larger voltage. Voltage is measured and converted to temperature.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Orifice plate

A

Fluid squeezes through a small hole which causes its velocity to increase. Pressure drop occurs after the hole, following Bernoulli’s Principle. Pressure difference measured by manometer or DP Cell. Pressure difference is proportional to the flow rate squared

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Fast opening CV

A

Small change in opening gives a large change in flowrate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Control valve coefficient

A

Q = Cv (ΔP / Sg)^(1/2)

  • Q = Volumetric flowrate through fully open valve
  • Cv = CV coefficient
  • ΔP = Pressure drop across valve

-Sg = Specific gravity

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Open loop control

A
  • Controller doesn’t monitor control variable
  • System operates on a pre defined set of instructions
  • Disturbances are not corrected
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Static gain

A

How much the output changes in response to a change in input

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Dead time

A

Distance / Velocity lag

Waiting period between when an action is applied and when it is first detected in the system. Hides disturbance from the process

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Capacitance

A

A system’s ability to store mass or energy. Acts as a natural buffer for disturbances

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

On/Off control mode

A
  • FCE is either on/off
    ADVANTAGE: Good for large capacitance systems where the PV changes slowly when the FCE is turned on/off

DISADVANTAGE: FCE gets worn out quickly and cycling can occur (repeats between on/off)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Large K_c (proportional gain)

A
  • Reduces offset
  • Can introduce instability as small changes in error will lead to large control action, causing system to overcompensate and result in another error in the other direction, causing oscillations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Proportional control

A

mv = Kc * e
e = SP - CV

ADVANTAGES: Simple and easy to tune (only Kc is needed)

DISADVANTAGES: Cannot remove offset as error is needed to obtain value for MV

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

PI control

A

mv = Kc ( e(t) + 1/t1 * integral[e(t) dt])

ADVANTAGES: Integral term adds up error over time, causes mv to change even with constant error (offset), this causes the error to gradually reduce with time until the error is 0. At this point, the integral term is equal to the sum of all the past errors and mv is constant. Eliminates offset

DISADVANTAGE: Longer response period

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

PID control

A

Derivative term uses the rate of change of the CV to determine how fast the output is changing and adjusts the value of MV accordingly to ensure it doesn’t overshoot the set point (create another error).

DISADVANTAGE: Does not work well in the presence of noise (CV jumps around alot)

Tuning is harder as more parameter must be specified

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Programmable logic controller

A

Industrial computer designed to automate processes and machinery

ADVANTAGES: Fast, rugged
DISADVANTAGE: Difficult with complex processes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Distributed control system
Use of multiple computers spread across facility to control different processes automatically. Very expensive
25
Overdamped response
Slow response with no oscillation from SP
26
Critically damped
Fastest response with no oscillation
27
Under damped
Fast return to SP with lots of oscillation
28
Integrated absolute error
Calculates total error for set point (including positive and negative deviations). Used to compare different tuning parameters.
29
Integrated square error
Punishes larger deviations more heavily. Used when minimal deviation is not bad
30
Integrated Time Absolute Error
Punishes errors that persist over time. Used for systems that need quick recovery
31
Trial and Error tuning
- Start with P, I, D all at a minimum - Adjust Kc by factor of 2 - Judge against if it meets ur standards of criterion - If not, decrease integral and derivative action to minimum and readjust
32
Reaction curve tuning
- Process is allowed to reach steady state - Do not have any control action and introduce a small disturbance - Observe how PV changes Apply a step change and record the process variable's response over time. This gives u a reaction curve
33
Reaction curve tuning parameters
L = Lag time (min) - Time before PV responds to change T = Time constant estimate (min) - Time for PV to reach ~63% of its final change P = Initial step disturbance (%) - Size of step change applied to process input deltaCp = Change in PV in response to step change N = Reaction rate - Speed of PV response (deltaCP/T) R = Lag ratio - How sluggish is response (L/T)
34
Constant cycling tuning (ziegler nichols)
- Attach P only controller with no gain - Place controller in automatic (it autonomously adjusts output) - Introduce small disturbance and gradually increase Kp until continuous cycling occurs - Find Pu (Ultimate period) = Time for one full oscillation - Find Ku (Ultimate gain) = The gain that causes the continuous cycling
35
Cascade control
Use of two controllers with one feedback loop nested within the other. The output of the primary controller used as the set point of the secondary controller which manipulates the FCE. Used when manipulated variable experiences disturbances
36
Feedforward control
Type of control that eliminates disturbances before they occur. Measures disturbance before it propagates in the system. Uses predefined math equations to see how it will affect the process. Adjusts the FCE accordingly to eliminate the effect of that disturbance Used if all possible disturbances are known (impossible)
37
Ratio control
Type of control that keeps the ratio between two variables fixed. One is 'wild' and 'uncontrolled' while the other is manipulated. Both flowrates are measured and ratio calculated and compared to set point ratio
38
Override control
Used when number of variables required to be controlled exceeds number of manipulated variables (T and P in batch reactor). Measure each control variable and compare to set point using controller. The output signal from each controller is then sent to the FCE (each one tells the valve a different amount to close by). Selectors are used to choose the actual amount the valve closes by. High selector chooses the highest controller signal to be sent to the FCE while the Low selector chooses the lowest signal.
39
Auctioneering
When multiple temperature transmitters are used in one system. They each pass through a series of high selectors until the highest temperature is chosen. This is the measured variable sent to the controller, which compares to the set point and adjusts the FCE accordingly
40
Split range control
Used when there is 1 CV and multiple MVs such as when both hot and cold water is used to control reactor temperature
41
Flow + Liquid Pressure loops
- Fast response (no capacitance) - Output flowrate only limited to how fast valve moves - Noisy so use PI control
42
Liquid level loop
- Large capacitance - Noisy due to turbulence and stirring - Consider P initially, then use PI
43
Gas pressure contorl
- Use P only - Self regulating nature means loop is always stable
44
Temperature control using Level
- Use PI or PID because temperature measurement is not noisy
45
Relative gain array
Quantitative method of analysing the level of interactions between loops, providing a method of pairing MV and CVs. DISADVANTAGE: Only checks variable during Steady state, not dynamic situations where it may be too slow to respond
46
RGA equation
Change in CV with change in MV when all loops are open / Change in CV with change in MV when all other loops are closed and working perfectly
47
Relative gain = 1
Other control loops have no interaction
48
Relative gain = 0
MV has no effect on CV
49
Relative gain > 1
Open loop gain is greater than closed loop gain. Other loops reduce effectiveness of chosen input/output
50
Relative gain < 0
Other loops invert control action, causing instability as MV will continue to rise/decrease to correct inverted change by other loops
51
Variation management
Short term: Transfer variation to less critical variation (e.g. transferring product purity variation to level/pressure) Long term: Changes to process to eliminate variation (e.g. add filter on output stream to increase product purity)
52
Plant control heuristics
- Control F and P once - Only 1 control valve per line - Control level wherever there is L/V interface - Minimise T and composition control
53
Model predictive control
Computer control scheme that uses a process model to predict future plant behaviour and the correct control action to drive CV to the SP.
54
MPC advantages
- Good for multivariable processes - Handles dynamic systems well - Optimises control variables according to constraints on the system
55
Elements of MPC
- Reference trajectory = Desired path for future outputs - Process variable prediction = What the process variable will actually behave like based on the MV or current control action - Error prediction update = The difference between how you want the output to behave and how its currently behaving
56
Fault
Deviation of at least one variable from acceptable behaviour
57
Failure
Permanent interruption of a system's ability to perform a desired function under operating conditions
58
Fault detection methods
MODEL BASED: Uses a mathematical model of the process and compares sensor results with estimated results from model SENSOR BASED: Analyses raw sensor signals to check for abnormal patterns indicating a fault KNOWLEDGE BASED: Uses historical data to compare with observed behaviour
59
Advantages of Model-Based Fault Detection:
- Early detection of faults - minimal sensors required - fault isolation - Adaptability
60
Assumptions for balance equations in a tank system
- Tank is perfectly mixed - KE and PE negligible - Specific enthalpy is equal to specific internal energy - Enthalpy only a function of temperature - Density is constant - Heat capacity is constant - No shaft work
61
Steady state gain
Change in the deviation variable when the step change in the forcing function is equal to 1
62
Time constant
Time taken for system to reach new steady state
63
List the three main physical situations in which a second order system may arise.
1. Multicapacity process 2. Inherently second order systems 3. Process system including its controller

Process dynamics (1 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions