Thunderstorm Characteristics of Importance to Wind Engineering

1
Thunderstorm Characteristics of Importance to
Wind Engineering

• Franklin T. Lombardo, Ph.D.
• Texas Tech University
• Lubbock Severe Weather Conference
• Lubbock, Texas
• February 18, 2010

2
PROBLEM STATEMENT

• Wind is Wind
• Statistics for wind/pressure used in wind load
  standard (ASCE 7)
• Wind Tunnel Data ? steady mean and variance ?
  stationary (log-law)
• Validated with full-scale data that is stationary
  in boundary layer (SBL) over periods ranging from
  10 minutes to 1 hour (spectral gap)
• Extreme events (e.g. thunderstorms, hurricanes)
  --gt non-stationary ? control design in most of
  the US
• Assume that physical and statistical
  characteristics are the same

An example of a stationary wind record (left) and
a thunderstorm record (right)
3
INTRODUCTION

• Non-Stationary Wind/Pressure Data
• Wind/Pressure Statistics (e.g. turbulence
  intensity, pressure coefficient)
• Use mean wind speeds within the spectral gap
• Thunderstorm usually occur over durations shorter
  than the spectral gap ( 1-10 min) and display
  non-stationary characteristics, especially short
  duration ramp-up events
• Difficult to make comparisons between stationary
  and non-stationary data statistics not
  representative
• Attempt to collect additional thunderstorm data
  and facilitate comparisons of the two events

4
INTRODUCTION

• Facilities/Instrumentation/Data Collection
• Wind Engineering Research Field Laboratory
  (WERFL)
• 200 Meter Tower
• Meteorological instrumentation on 10 different
  levels ? 3 to 656

• 204 differential pressure taps (building) (104
  walls, 90 roof)
• 30 sonic ? geometric center
• 160 tower ? 5 levels
• 150 feet away
• Now at Reese ? building remains
• 13 Tower, 30 Sonic

5
THUNDERSTORM EVENTS

• Ramp-Up Types/Characteristics
• Exhibit rapid increase/decrease in wind speed
  over a short period
• Time histories show some similarities but no
  universal form (wide variability)
• Some occur over longer scales ( 2 min), others
  shorter ( 10 sec) ? 9 events

58
6
THUNDERSTORM TIME SCALES

• Andrews AFB Microburst (1983) ? 90-100 seconds
• Standard for wind engineering use 150 mph gust
  poor data quality
• Lubbock RFD (2002) ? 100 seconds (Holmes, 2008),
  2 3 minutes (Kwon and Kareem, 2009)
• 90 mph gust ? design wind speed for most of the
  country high resolution data
• Want to determine information of importance to
  wind engineering
• Previous studies used time-varying mean for
  non-stationary events to quantify information
• Created algorithm to measure durations of
  stationary turbulence
• Stationary turbulence that contained peak wind
  speed was used

7
RESIDUAL TURBULENCE
Using 17 second averaging time Mean Residual
Turbulence Duration 150 s Appropriate time
periods for analysis in thunderstorm prone areas
should be 60 200 seconds These representations
(using 15 60 s averaging time) can be used for
further wind engineering statistics (TI, GF,
PSD) Likely areas a higher turbulence on small
scales shown in previous figure (10s) but would
be near impossible to quantify
8
THUNDERSTORM VARIABILITY

• So what does the reduced time scale and
  consideration for thunderstorms in structural
  design mean?
• Increased Variability
• Other studies (Ponte and Riera, 2007) have shown
  highly varying time scales for thunderstorms
• Other variability has been shown in vertical wind
  speed profiles, turbulence, etc ? will show
  later
• Assuming statistical and physical properties are
  the same for a moment

100 s
900 s
Schroeder (1999)
9
WIND ENGINEERING PARAMETERS

• Turbulence Intensity
• Compared with SBL data (100 s segments)
• All ramp-up events fall within range of SBL
  (33) for 15-60 s averaging times
• VORTEX2 case outside of range gt 10 second
  averaging time (7 )
• Inherently additional turbulence, but likely not
  attributed to surface roughness

10
VORTEX2 CASE

• May 15, 2009 ? North Central Oklahoma
• Although wind speeds barely exceeded severe
  levels and are well below design values for a
  short period, it raises a number of interesting
  questions for wind engineering as it is a unique
  time history (TI values different)
• Multiple rapid changes in wind speed and
  direction 2 minute period
• Periodic fluctuations on relatively smaller
  scales (0.03 0.05 Hz)
• Also small spatial scale ? probe 1 mile away
  did not record event

11
WIND ENGINEERING PARAMETERS

• Gust Factor
• VORTEX2 case, others, outside of range gt 100
  seconds, smaller time scales
• Higher variability noted, few straddle bounds of
  SBL although most within
• Suggests similar gustiness at short time scales
• Ramp-Up GF different than one used in ASCE
  60-100 seconds
• V2 GF for a 1500 second record was 9

12
WIND ENGINEERING PARAMETERS

• Power Spectral Density, Turbulence Scales
• Look at turbulence in frequency domain high
  frequency scales (along-wind component)
• At frequencies gt 0.05 Hz, thunderstorm energy is
  similar to SBL models
• However V2 case shows strong energy at 0.03-0.05
  Hz (not shown)
• Other cases show strong energy at 0.01 Hz

13
VERTICAL WIND PROFILES

• Important for Structural Loading
• ASCE 7 assumes modified log profile for 3
  second gust wind speed
• Evolutionary factors not considered in wind
  engineering
• Design exceedance at only one or multiple levels
• Taken from 200 meter tower ? Reese Field Site

• Transition from SBL to impinging jet ? 30s
• Momentum works downward with time
• Below maximum wind speed ? resembles SBL
  profiles (low as 13)

14
VERTICAL WIND PROFILES

• Other Examples
• Some cases show close to uniform profile noted
  in other extreme wind studies
• Compared with SBL 3-second maximum gust profiles
• 0.30 z/zmax compared to 0.88 z/zmax for SBL
  (highly variable)
• Environmental conditions, storm type (i.e.
  isolated microburst, bow echo, supercell) need to
  be further studied
• Highest wind speed at surface similar whereas
  highest overall wind speed from HP supercell/bow
  echo

June 19, 2008
June 4, 2009
June 19, 2003
Impinging Jet
Log
Uniform
15
VERTICAL ANGLE OF ATTACK (33)

• Noted in studies (Wu, 2001 Richards and Hoxey,
  2004) to induce high negative pressures on roof
  with positive (upward) anglesNOT vertical wind
  speed
• No significant differences detected versus SBL
• Even in ramp-up events due to strong horizontal
  wind speeds
• May be different as surface roughness becomes
  less dominant
• Strong upward motion in tornadic vortices, for
  high-rise buildings gt 60 feet

16
BUILDING EFFECTS

• Pressure Coefficient vs. Angle of Attack (3
  second)
• Use sonic (30) on top of WERFL assuming (2
  events)
• Uniform profile, no angle of attack changes from
  MRH to 30
• Use (13) 150-200 from WERFL (1 event)
• Determine any flow field differences over that
  distance

?
17
BUILDING EFFECTS

• 95 of ramp-up Cps (red) fell within range
  of WERFL SBL at similar AOA using peak 3-s gust
• All fell within range in conical vortex regions
• Flow features over building are similar

18
BUILDING EFFECTS

• Interest of what happens in separation region
  during gusting conditions (Murgai et al.,
  2006Hwang et al., 2001)
• Temporal acceleration of wind has become area of
  interest (Doswell et al., 2009)
• Criteria 20 mph increase in 3s, flow normal to
  walls (gust, mean), AOA constant

• Determination of
• Distance of Strongest Negative Pressure From Roof
  Edge
• Aerodynamics Changes

19
BUILDING EFFECTS

• Results
• Mean cases ? 3.9 4.1 feet
• Gust cases ? 2.0 5.3 feet ? high variability
• Pressure distributions similar when using mean
  gust speed
• Anemometer 30 feet away ? still difficult to
  determine the effects at smaller time/length
  scales ? correlation of wind and pressure

• May actually be gain additional information in
  wind tunnel where wind/pressure effects can be
  more easily measured/visualized

20
EXTREME WIND SPEED ANALYSIS

• Current ASCE wind map uses basic wind speeds
  (3s gust) without regard for storm type and
  assumed uniform exposure
• Computation of design pressure on a building for
  all US (most 90 mph)
• Thunderstorm winds shown to have different
  probability distributions and dominate most US
  extreme wind climates including West Texas
• 200 ASOS stations in current analysis high
  resolution data (WTM, StickNet), additional ASOS
  available to enhance current wind estimates (6
  exceedances in 8 years) GIS programs to aid with
  address roughness issues
• Due to small spatial scales (V2, others), wind
  speeds not in current analysis

21
EF-SCALE ISSUES/QUESTIONS

• Main application is tornadoes but these research
  topics would apply to thunderstorm research as
  well
• Temporal/Spatial character of high winds
• Temporal Acceleration
• Duration vs. Damage Flow Modification
• Coherence/Correlation
• Wind Speed vs. Damage Relation
• Rapid Wind Direction Changes ? affect building
  pressures
• Additional high resolution measurements
• StickNet, KA Band Radar ? near surface wind
  characteristics
• Pressure measurements on structures similar to
  hurricanes
• Vertical wind speeds in tornadoes
• Does it offset the strong horizontal wind speeds?

22
CONCLUSIONS/FUTURE WORK

• Extreme thunderstorm events (9) studied for wind
  engineering purposes
• High Variability (time series, time scales, WE
  parameters, vertical profiles)
• Time Scales ( 60 -200 seconds)
• Current method not appropriate for analysis in
  thunderstorm areas
• Likely small scale turbulence regimes not
  accounted for
• Wind Engineering Parameters
• Turbulence Intensity ? SBL, TS similar for
  prescribed averaging times with exception of V2
  case
• Gust Factor ? high variability, gt 60 seconds no
  Durst Curve
• Power Spectral Density ? periodic fluctuations
  evident, higher scale turbulence important to
  most structures similar
• Events like V2 case need additional documentation
  and study
• Vertical Profiles
• Evolve over short time scales maximum profiles
  highly variable
• Peak on average lower the max measuring height

23
CONCLUSIONS/FUTURE WORK

• Extreme thunderstorm events (9) studied for wind
  engineering purposes
• Vertical Angle of Attack
• No significant differences compared to SBL
• May be different at higher above surface,
  tornadic cases
• Building Effects
• 3-s Cp mostly within range of SBL ? all in
  critical areas
• Rapid increases in wind speed do not seem to
  alter aerodynamics
• Rapid wind direction changes need study
• Extreme Wind Speeds
• Can be further enhanced with field programs to
  capture events of small temporal, spatial scales

24
QUESTIONS/COMMENTS