Instrumenting, Post-tensioning, Recording and Modeling the Moose Brook Howe TrussesModeling the Moose Brook Howe Trusses

by

Gasparini, D., Nizamiev, K., Harnar, N. and Berilla, J. p , , , , , ,Department of Civil EngineeringCase Western Reserve University

Cleveland OHCleveland, OH

Wood

Anisotropic – Longitudinal, radial and tangential directionsInhomogeneous – Mechanical properties vary from point to pointViscous - Stress-strain behavior is a function of timeViscous Stress strain behavior is a function of timeHygroscopic – Ingress/egress of moistureFlaws – Knots, checks, splits, etc.

Highly variable mechanical properties that are a function of wood moisture content

Atmospheric conditions and truss behaviorp

Temperature and relative humidity vary continuously in time

Thermal strains, viscous behavior, and shrinkage/swelling strains from moisture ingress/egress are “concurrent” in

i i Astructural wood components in an outdoor environment. All three phenomena affect forces in post-tensioned Howe trusses.

Scope of work and objectives

•Instrument Moose Brook Howe trusses with strain, temperature and relative humidity sensorsy

• Post-tension the trusses and place them outdoors (but protected)• Acquire and interpret data from truss and atmospheric sensors

for a period of at least one yearfor a period of at least one year• Mathematically model the trusses and assess models• Make judgments on structural significance of viscosity,

th l d h i b h ithermal response and hygroscopic behavior

• Improve understanding of structural behavior of Howe trusses• Improve rehabilitation technologies for Howe bridges• Improve designs of new Howe bridges

Weldable full bridge strain gauges, HPI model HBWF-3 12 6 10G 1” C35-125-6-10GP-TR-1” PC applied on Diwidag bars

Installation of Sensirion SHT25 semiconductor T/RH sensorsInstallation of Sensirion SHT25 semiconductor T/RH sensors

4 mm

Instrumentation of two Howe trusses

InstrumentationSingle 8x8 post-tensioned specimeng p p2 Steel stain gauges on two stressed bars1 Steel strain gauge on unstressed bar section1 Atmospheric T/RH sensor1 Atmospheric T/RH sensor1 T/RH sensor in small loose wood block3 T/RH sensors at three depths within the cross-section

Two Howe trusses18 Steel strain gaugesg g1 Steel strain gauge on unstressed bar section1 Atmospheric T/RH sensor14 T/RH sensors in various members at various depths14 T/RH sensors in various members at various depths

Data acquisition systemD i d b ilt d it d b Ji B illDesigned, built, and monitored by Jim Berilla.Jim.Berilla@case.edu

Post-tensioning of single element

Si l 8” 8” D l fiSingle 8” x 8” Douglas fir post-tensioned specimen

Post-tensioning of two Howe trusses

Architectural fabric cover designed, fabricated, and supplied pro-bono by the Seaman Corp. of Wooster, OH

8x8 specimenp

Post-tensioned on 4-3-2012; installed outdoors on 5-8-2012

Howe trusses

i 6 13 2012 i 9 2 2012Post-tensioned on 6-13-2012; installed outdoors on 9-25-2012

F i S t 1 f d t

104

106Fourier Spectrum: 1 year of data

365 sinusoidal cycles (day)1 sinusoidal cycle (year)TemperatureForce 8x8 specimen

102

Force 8x8 specimenRelative Humidity

100

10-4

10-2

10-8 10-7 10-6 10-5 10-410-6

Frequency (Hz)

q y ( )

8x8 specimen – Fourier amplitude spectra of recorded time histories

Viscous losses likely from stresses normal to the grain in the chords; minimize by using castingsminimize by using castings with “sleeves”

Mathematical modelingMathematical modeling

Linear viscoelastic analyses Three parameter solid modelBurger model

Diffusion analysesTwo-dimensional isotropic diffusion model

Three-parameter solid model for wood; linear elastic model for steel

Four-parameter Burger model for wood; linear elastic model for steel

Effect of Temperature increase (T = - 40 degrees )

50

50.5p ( g )

3 parameter (Solid) model + Steel +Temp4 parameter (Burger) model +Steel+Temp

49.5

el (K

ips)

48.5

49

ce in

Ste

48

Forc

0 10 33 50 75 10047.5

Days

52Combined Effect of Various Axial Stiffnesses without drop in Temperature

50

52

Kip

s ) 0.625 0.875 1.25

Prestress Condition after all immediate elastic deformation occured

48

g B

ars

( Prestress Condition after all immediate elastic deformation occured

8.6 % loss

44

46

n D

ywid

ag %

12.9 %

42

Forc

e in

17.7 %

0 10 20 30 40 50 60 70 80 90 10040

Days

Four Parameter Burger Model: Influence of the ratio c = 1 / 2

49

501 2

c = 100 c = 1000 c = 10000

48

( Kip

s )

47

in S

teel

45

46

Forc

e

0 20 40 60 80 100 120 140 160 18044

d

days

Diffusion model for the egress/ingress of moisture

2 2

2 2( )C C CDx y t

Two Dimensional Isotropic

Fickian Diffusion Modelff

D Diffusion coefficientC Concentration

( ( ))CD S C C t /Emission Admission Model( ( ))b eD S C C tx

/Emission Admission Model

at boundary

( )e

S Surface emission coefficientC t Concentration in equilibrium with atmospheric conditionsC C t ti t b d

bC Concentration at boundary

Initial conditions and boundary conditions

7.25 inch

RH/Temperature Sensorsat 3 locations to capture EMC gradient within the depth ofth 8 8 i

7 25 inch

the 8x8 specimen

7.25 inch

1 82 i h

1 inch

3.63 inch

1.82 inch

Assumed initial moisture content distribution within cross-section

D = 1.12e-2, Influence of the scale

18.5

19D 1.12e 2, Influence of the scale

EMC center, specimen size: 12x12 inchEMC boundary, specimen size: 12x12 inchEMC t i i 7 25 7 25 i h

17.5

18 EMC center, specimen size: 7.25x7.25 inchEMC quarter, specimen size: 12x12 inchEMC boundary, specimen size: 7.25x7.25 inchEMC quarter, specimen size: 7.25x7.25 inch

16.5

17

EM

C %

15.5

16

0 50 100 150 200 25014.5

15

d

days

Size of the specimen: 7.25x7.25 inch, Influence of the Diffusion Coefficient

18.5

19Size of the specimen: 7.25x7.25 inch, Influence of the Diffusion Coefficient

EMC center, D = 0.56e-2EMC bound, D = 2.24e-2EMC quarter, D = 0.56e-2

17.5

18q ,

EMC center, D = 2.24e-2EMC quarter, D = 2.24e-2EMC bound, D = 1.12e-2EMC center, D = 1.12e-2

16.5

17

EM

C %

,EMC quarter, D = 1.12e-2EMC bound, D =0.56e-2

15.5

16

0 50 100 15014.5

15

d

days

Influence of the surface emission coefficient on moisture content within cross-section

D = 0 2e 2 Specimen Size: 7 25x7 25 period 365 days S = 6 86e 1

18

19D = 0.2e-2 , Specimen Size: 7.25x7.25 , period 365 days , S = 6.86e-1

16

17Experimentally  measured

14

15

EMC

%

13

14

EMC center (predicted)

0 50 100 150 200 250 300 35011

12

D

EMC center (predicted)EMC shallow (predicted)EMC quarter (predicted)

DaysPredicted vs.  measured moisture content  time histories  with the observed actual atmospheric temperature and relative humidity fluctuations as boundary conditions

ObservationsTh 8 8 i l t i t l 8% f it t•The 8x8 specimen lost approximately 8% of its prestress

over one year due to wood viscosity; the present loss rate is extremely small.y

• The prestress losses in the Howe trusses are approximately 30% and 50% but these could be minimized by use of30% and 50%, but these could be minimized by use of “sleeved” nodal castings.

It’ l t t i th t ith t t i• It’s almost certain that with current prestressing processes a permanent state of pre-stress can be achieved in Howe trusses.

• Atmospheric T and RH fluctuations do affect forces in Howe trusses, but the stress ranges in the wood and steelHowe trusses, but the stress ranges in the wood and steel are small.

ObservationsObservations

• Effects of temperature variations are predicted well by li l ti d llinear elastic models

• Diffusion models can be calibrated to predict wood moisture content variations that match observed values well.

Additional work

C ti i i d t th h th d f 2013• Continue acquiring data through the end of 2013

• One of the Howe trusses will be re-tightened to observe changes in subsequent viscous behavior in wood

• A coupled “mechanosorptive” model is required to A coupled mechanosorptive model is required to predict stresses from wood strain caused by moisture content variation; the diffusion model predictions may be h l f l i d fi i ff ti i l t i f i thelpful in defining an effective axial strain from moisture ingress/egress.

Suggestions……

If a new covered bridge is commissioned,Select a Howe truss in its original formUse nodal castings with “sleeves”Use nodal castings with sleevesUse low moisture content woodNo need for bolts, gusset plates, adhesives, fiber-

i f d l l dreinforced wraps, trunnels, complex wood joinery, etc.

Maximize shop fabrication and pre-assemble trussesp pUse current post-tensioning technologiesPost-tension in summerAchieve a permanently post-tensioned wood Howe bridge!Achieve a permanently post-tensioned wood Howe bridge!

If a Howe bridge is to be rehabilitated,D ’t ll / l k t di lDon’t allow/assume slack counter-diagonalsControl initial tightening and prescribe re-tightening

A k l d tAcknowledgments

FHWA/NHCBP – Sheila DuwadiNPS/HAER – Christopher MarstonNSPCB – David WrightBBNE – Tim AndrewsBBNE Tim AndrewsCWRU team – Kamil Nizamiev, Neil Harnar, Jim BerillaSeaman Corp. – Ken Chaloupek, Scott Burgess