PULL THROUGH FAILURE

PERFORMANCE OF LIGHT

GAUGE STEEL ROOF BATTENS

UNDER SIMULATED CYCLONIC

WIND LOADING

Daniel Mill and Sudarshan Bhandari

Cyclone Tracy (1974)

Darwin Area Building Manual (DABM)

Northern territory implemented the DABM in 1975

Consists of 10000 cycles of permissible load to 1.8 times of that load until it

fails.

TR440

DABM has considered to be too conservative and TR440 been introduced at

1978

Consists of simple loading sequence programme of

Load Cycles

62.5% of design load (DL) 8000

75% of DL 2000

100% 200

gxDL (g= no of tests) overload

Low High Low (LHL)

Since the actual imitation of cyclones have not been considered by TR440,

the more likely (similar to cyclones) loading sequence have been proposed by

Mahen Mahendran on 1994, also listed in NCC.

Sequence Number of cycles

A 4500

B 600

C 80

D 1

E 80

F 600

G 4500

Basic failure modes

Two basic types of failure mechanism found on roof battens to the rafter

connection.

Pull-out failure

Pull-through failure

Research Problem Definition

Accurate prediction on capacity of the battens are necessary.

The current equation listed in the AS4600 overestimates the pull through capacity and gives a lot higher values than the actual experiments.

Static tests (slow rate loading) has been done in previous tests to accurately predict the actual pull through capacity.

Same capacity have been applied to the LHL tests to predict the cyclic capacity.

Some researches found that the cyclic capacity is actually higher than static capacity which gives a very unusual results.

Static capacity is taken when the crack initiates and load starts to drop, however cyclic capacity is taken when the member disengage due to complete pull through.

Research Aim

Identify past research and testing relating to the low-cycle fatigue effect on

roof batten connections

Test not only 0.75mm roof battens but also 0.55mm to further enhance results

and explore the possible effect of thickness on roof batten fatigue capacity

Results to provide complete pull through failure capacities of the 0.75mm and

0.55mm battens for multilevel-level cyclic tests

If results confirm static capacity is still lower than cyclic capacity, provide

recommendations.

Assess the impact of loading rate on the static capacity of roof battens

Specimen details (0.55mm and 0.75mm)

Batten testing regime

0.55mm thick two span slow load rate static tests

0.75mm thick two span slow rate static tests

0.55mm thick cantilever high load rate static tests

0.75mm thick cantilever high load rate static tests

LHL

Testing setup- Static two span and LHL

Testing setup - Cantilever

Test plan

Test 0.55mm Batten 0.75mm Batten

High Speed Pull-Through

Cantilever (Static)

3 tests 3 tests

Slow Speed Pull-Through

Cantilever (Static)

1 test 1 test

Two-Span Test (Static) 3 tests 2 tests

Multi-Level Cyclic Test 1 tests 2 tests

ResultsTwo span slow loading rate

0.55mm static test (slow loading rate) 0.75mm static test (slow loading rate)

High loading rate cantilever

0.55mm static test (high loading rate)0.75mm static test (high loading rate)

Two span static (slow rate loading)

Batten Type Static Capacity

(kN)

Mean COV

0.55mm (TS4055) 2.39, 2.29, 2.27 2.32 0.02

0.75mm (TS4075) 3.10, 3.10, 3.12 3.11 0.00

Multi-level Cyclic tests0.75mm batten with 100% slow rate loading rate static capacity

Multi-level Cyclic tests0.75mm batten with 110% slow rate loading rate static capacity

Multi-level Cyclic tests0.55mm batten with 100% slow rate loading rate static capacity

Discussions and analysis

Effect of Loading Rate on Static Capacity of Roof Battens

Test Batten Size

High Load

Rate Static Capacity (kN)

Slow Load

Rate Static

Capacity (kN)

Cantilever 0.75 3.83 3.25

0.55 2.37 2.07

Discussions and analysis

LHL loading sequence seems to be too soft due to possible underestimation of

the higher loads in the middle cycles and the minimum holding criteria

defined by NCC at the peak cycle.

Just the initiation of crack for static failure criteria compared to complete

pull through for cyclic failure has not been equally accounted.

Results shows that the thinner members are more vulnerable to the cyclic

loading than thicker members, furthermore thinner members have lesser

effect due to faster loading rate.

Acknowledgements

Mahen Mahendran

Myuran Kathekeyan

Staffs at Banyo Pilot Precinct

QUT Library services

The end

Thank you