APMP TC Initiative Project - Research on the Calibration of 3D Pitot Tubes and

Flow Measurements of Greenhouse Gas Emissions

Participants:

CMS: Hsin-Hung Lee, Chun-Min Su

NMI: Liang Zhang, Chi Wang

KRISS: Woong Kang, Yong-Moon Choi

NIST: Iosif I. Shinder

September 2014

Speaker: Chun-Min Su

Contents

• Introduction

• Objective and collaboration

• Activities between project members

• Current progress

• Future work and planning

Where on earth are you most likely to die early from air pollution?

Introduction

Stationary Source Emissions in air quality terminology is any fixed emitter of air

pollutants, such as fossil fuel burning power plants,

petroleum refineries, petrochemical plants, food

processing plants and other heavy industrial

sources

Objective and Collaboration

Pitot tube characterization

(CMS, NIM, KRISS, NIST)

• L-type, S-type, Prism-type, Omni, Cobra

• CFD, Smoke visualization

Standard traceability (CMS)

• Differential pressure

Calibration method and facility

(CMS, KRISS, NIST)

• Nulling & Non-nulling method

• Traversing stage design

Uncertainty evaluation (KRISS)

• Stack flow

Calibration of 3D

Pitot Tubes and

Flow Measurements

of Greenhouse Gas

Emissions

Activities between Project Members

• Guest research at NIST (Dr. Hsin-Hung Lee, CMS)

– November 2012 to June 2013

• Guest research at NIST (Dr. Liang Zhang, NIM)

– June 2013 to June 2014

• Guest research at NIST (Dr. Woong Kang, KRISS)

– March to September 2014

• Project and progress discussion at NIST (Hsin-Hung

Lee, Liang Zhang, Woong Kang, Iosif I. Shinder)

– April 8th to 12th, 2014

Current progress (1) - Omni-type probe and pressure scanner

Omni type of 3D pitot tube Pressure scanner

𝐶𝑃_𝑃𝑖𝑡𝑐ℎ = (𝑃3 − 𝑃2)/(𝑃1 − 𝑃 ) (1)

𝐶𝑃_𝑌𝑎𝑤 = (𝑃5 − 𝑃4)/(𝑃1 − 𝑃 ) (2)

𝐶𝑝𝑇𝑜𝑡𝑎𝑙 = (𝑃1 − 𝑃𝑇𝑜𝑡𝑎𝑙 )/(𝑃1 − 𝑃 ) (3)

𝐶𝑝𝑆𝑡𝑎𝑡𝑖𝑐 = (𝑃 − 𝑃𝑆𝑡𝑎𝑡𝑖𝑐 )/(𝑃1 − 𝑃 ) (4)

𝑃 = 𝑃2 + 𝑃3 + 𝑃4 + 𝑃5 4 (6)

Piston-type Pressure Generator (CMS)

10 kPa 差壓 600 Pa 差壓

1 2 3 4

USB HUB

MOXA

10 kPa 600 Pa

活塞壓力

產生器

被校件 查核件 標準件 被校件 查核件 標準件

Range of differential pressure / Extended uncertainty

(0 to 600) Pa / U = 2.5 Pa

Current progress (1) - Pressure calibration

Current progress (1) - Calibration of pitot tube and traversing stage design

Pitch angle vs. Pitch angle pressure coefficient

Pitch angle vs. Velocity pressure coefficient

Pitch angle vs. Total pressure coefficient

Step 1: Align the probe so that the center hole is pointing towards a reference position. Step 2: Rotate probe until P2=P3. This is the Yaw angle. Step 3: Calculate Pitch Angle Pressure Coefficient [(P4-P5)/(P1-P23)]. Step 4: Determine Pitch Angle. Step 5: Determine Velocity Pressure Coefficient [(Pt-Ps)/(P1-P23)]. Step 6: Calculate Velocity pressure (Pt-Ps). Step 7: Determine Total Pressure Coefficient [(P1-Pt)/(Pt-Ps)]. Step 8: Calculate (P1-Pt) and obtain Pt.

Step 3

Step 4

Step 4

Step 5

Step 4

Step 7

Current progress (1) - Comparison of nulling and non-nulling calibration

Adaptive-Network-based Fuzzy Inference System (ANFIS)

0

2

4

6

8

Pitc

h

Yaw

CpA

CpB

CpB

C

pA

CpB

Y

aw

CpB

P

itch

CpB

P

itch Y

aw

CpB

P

itch C

pA

CpA

C

pB

P

itch Y

aw

Training (Circles) and Checking (Asterisks) Errors

RM

SE

ANFIS sequence forward search

Meshing

Pitot tube simulation data is very sensitive to the boundary layer

mesh

Ansys Meshing patch confine method is not consistent

Tetrahedral mesh fit experimental data better than Hexahedral

mesh

Finally the mesh was generated by the Octree method inside

Ansys Meshing(mesh independent and select ICEM File Output)

Current progress (2) - Meshing of pitot tube CFD simulation

Turbulent Models

k-e relizable, ke RNG, ke standard (different wall treatment)

K-w

K-w SST

RSM Linear Pressure Strain, RSM Quadratic Pressure Strain,

RSM Stress Omega

Transition SST

Transition SST and K-w SST give the best result compare

to the experimental data, same calibration factor,

Transition SST give more detailed flow filed

Current progress (2) - Turbulence model of pitot tube CFD simulation

Effects of drawing the total pressure hole and static pressure

hole

It is very important to draw the total hole

When the wind angle is zero, draw the static hole has no effect

Current progress (2) - Structure analysis of pitot tube by CFD simulation

Pitot tube mouth shape is important for calibration factor at

higher wind speed

It is better to get the exact mouth shape

Boundary condition test of wind

tunnel @10m/s

Stationary wall

Moving wall

No viscous wall

Pressure far field

After correction all give the same

calibration factor

Standard flowrate

measured

Current progress (2) - Boundary conditions of pitot tube CFD simulation

For the boundary layer of pitot tube, Y+ should be around 1

Current progress (2) - Boundary layer of pitot tube CFD simulation

Prism-shaped DAT type 3D Probe according to US EPA Method 2F Diameter = 1/4”, length = 20”, 5 holes for comparison with other types of 3D Pitot tubes Conduct calibration at NIST wind tunnel(August) and CMS wind tunnel(November) Investigate characteristics of yaw and pitch angle effects at KRISS wind tunnel Uncertainty analysis of calibration process and coefficients

NIST Wind Tunnel - Test section : 1.5m 1.2m up to 75 m/s

CMS Wind tunnel - Open type test section up to 25 m/s

KRISS Wind tunnel -Test section : 0.9m 0.9m up to 15 m/s

P1(Total Pressure) P3 P2 (Yaw angle)

P4 P5(Pitch angle) Y

X

Z

Flow

Pitch Yaw

Current progress (3) - 3-Dimensional probe for measuring stack gas velocity

Calibration procedure and ranges Nulling calibration procedure according to US EPA method 2F Pitch angle : -45° to 45 ° with interval 2.5 ° (Uncertainty 0.5 ° ) Yaw angle : nulling method (-4 ° to 4 ° ) Velocity : 5, 10, 15, 20, 30 m/s (Red= 2,100 to 12,600)

Pitch angle

Yaw angle

Traverse resolution = 0.1°

Current progress (3) - Calibration of Prism-Shaped DAT Probe at NIST

Experimental Set-up Calibrated NIST’s L-type Pitot tube as Reference velocity Simultaneous measurements in the uniform velocity profile area

Reference Pitot Tube

Prism-shaped 3D Probe

Temperature Sensor

Humidity sensor

Test Section 1.5 m 1.2 m

P1(Total Pressure)

P3 P2

(Yaw angle)

P4 P5(Pitch angle)

* Pressure Transducer Connections MKS A MKS B MKS C MKS D

Total (+) P1 P1 P1 P1

Ref (-) P2 P3 P4 P5

YOKOGAWA MKS E

Total (+) P1 P2

Ref (-) P2 P3

MKS A Total P1

MKS A Ref P2

Experimental Set-up 6 Pressure transducers for 5 holes and reference velocity

Manometer MKS 698A (1 Torr )

Signal conditioner MKS 670B

Manometer YOGOKAWA MT210(500 kPa)

Calibration Results Pitch angle calibration curve (versus pitch angle)

𝐹1 = 𝑃4 − 𝑃5

𝑃1 − 𝑃2

Calibration Results Velocity calibration curve (versus pitch angle)

𝐹2 = 𝐶𝑃

∆𝑃𝑠𝑡𝑑

𝑃1 − 𝑃2

Cp : Reference Pitot tube Coefficients ∆Pstd : Reference Pitot differential Pressure

Future work and planning

Pitot tube characterization

• L type pitot tube in different pitch and yaw angle

• Effect of static hole

• Cobra probe simulation

• Investigate characteristics of yaw and pitch angle

effects near prism-shaped probe

Standard traceability

• Integrated testing with pressure scanner and

traversing stage for 3D pitot tube calibration

Calibration method and facility

• Conduct calibration at KRISS wind tunnel and

CMS wind tunnel

• Comparison of calibration methods

Uncertainty evaluation

• Uncertainty analyses for calibration procedure and

coefficients at each system

• Uncertainty analyses for measuring volumetric flow

rate in stack by Prism-shaped probe

Calibration of 3D

Pitot Tubes and

Flow Measurements

of Greenhouse Gas Emissions

Thanks for Your Listening