A Simple Method For Measuring Local Buckling of Thin Plates

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

A Simple Method for Measuring Local

Buckling of Thin Plates

by Z.T. Sang, K.C. Chang and G.C. Lee

ABSTRACT--A simple computer-interfaced optical system for measuring out-of-plane deformation is the 'shadow moire'
measuring the dynamic-local-buckling deformation of thin- method, 2 which uses a transparent plastic board placed
walled metal structural-plate elements is described in this parallel to and in front of the test specimen. The reference
paper with two sets of experimental results. The major ad-
grid is put on the plastic board in advance. When a light
vantage of this system is its simplicity and economy as well as
its speedy automated process for data scanning, acquisition, source is projected from one side of the specimen (not
and analyses by using a microcomputer. perpendicular to), the out-of-plane deformation will
generate image patterns on the reference grid. These
images can be used to estimate the deformation of the
Introduction specimen.
The strength of thin-walled steel structures subjected to The approach reported here adopts the general principle
in-plane loading is often controlled by local buckling. For of both these methods. A video camera and a micro-
economic reasons local buckling of thin structural elements computer are also used so that data scanning, acquisition
is permitted such as in earthquake-resistant design. and analyses are integrated into an automatic system.
However, there remains much to be understood about
inelastic buckling behavior of thin-walled structures under The Measuring System
dynamic-loading conditions. This lack of information
may be attributed to the difficulties encountered in The system hardware consists of a microcomputer, a
designing a laboratory system that can acquire time- 35-mm slide projector, a video camera, and an interface
dependent data in a speedy and reliable fashion. This card as shown in Fig. 1. The functions of each individual
is the primary concern of the present paper. component are as follows.
One method widely used for measuring local-buckling (a) The slide projector provides intensive parallel light
deformation in the 1970s was the 'sterocamera' method, ~ lines on the white-coated surface of the specimen. This is
which uses two identical cameras placed in front of the done using a 35-mm slide which has parallel, transparent,
specimen on which square grid patterns are drawn in equal-distant lines prepared in advance. The intensity of
advance. During the test these two cameras take pictures the light lines is directly proportional to the accuracy of
continuously. The out-of-plane deformation of the spec- the results.
imens is then estimated from the changes in the. square- (b) With the microcomputer system, the pictorial
grid patterns. The deformation history is obtained by messages received from the video camera in RAM are
manually measuring the changes of the square grids of the stored in the floppy disks. During the test the picture
pictures after being processed. A second approach in received can be displayed on the monitor. After the test is
completed the messages can be retrieved from RAM or
the disks for data analysis. We have been using an Apple
II computer in the current system. Results of some
preliminary tests are given to illustrate the application of
the system. At present the system is updated using an
z.T. Sang is Visiting Scholar, Beijing Polytechnic University, Beijing, IBM PC for more serious local-buckling experiments. The
China. K.C. Chang is Assistant Professor, and G.C. Lee is Professor and updated system will be described in the last section of this
Dean, Department of Civil Engineering, State University of New York at
Buffalo, Buffalo, N Y 14260. paper. Because the clock of an Apple II computer is only
Paper was presented at the 1986 SEM Spring Conference on Experimental 1 mHz, Assembly language is used in order to save
Mechanics hem in New Orleans, LA on June 8-13. processing time. As such it is possible to scan 64 data or
Original manuscript submitted: August 19, 1986. Final manuscript more per picture under a time rate equivalent to that of a
received." April 27, 1987. strong earthquake-induced dynamic-loading condition.

20 9 March 1988
TEST SPECIMEN

' / ~ "%,.
",% ," "--o
Fig, 2--Out-of-plane deformation of
structural member measured by the
', /
,,.. current system

I
IINTERFACE
CARD
I I- - I

. 'i
AB
SLIDE PROJECTOR MICRO-COMPUTER
Fig. 1--System hardware consisting of a microcomputer,
a slide projector, a video camera and an interface card
UIOEO CANERA

(c) In choosing a solid-state video camera, there are


two general types of commonly used computerized-image
cameras: the Vidicon type and the CCD (charge-coupled LIGHT SOURCE
device) type. Both serve well in our measuring system. Fig. 3--After the system measuresthe
Because of their cost, however, we have tried and are in-plane movementsof light line a'b, the
satisfied with an optical RAM IS323-~ in this preliminary out-of-planedeformation ab can be
model. It is essentially a dynamic RAM chip, which also calculated
has light-sensing capacity; and can be used as the picture
sensor. In this 256 • 256 bit RAM each bit acts as an
individual light sensor with two states: '1' or '0'. A '1' bit
appears when light is sensed.
(d) A camera interface card is necessary for the BB. After the out-of-plane deformation (cross section
communication between the computer and the solid-state moves from A A to BB), the light point originally on b
video camera. We use a Micro-D-Dam in this current moves to a ' from the direction of the camera. If the
system. 6 The clock frequency of this interface card is distance between b and a ' can be measured, and the
only 150 kHz. However, for 64 data per picture the curvature change in the scanning direction is small, then
scanning speed can be as high as 0.013 s/picture. the out-of-plane deformation ab can be obtained from
The system software consists of a main program written simple geometrical consideration.
in Basic language. Certain data-acquisition and data- In our experiments there are eight real light lines
processing subroutines are written in Assembly language projected on the surface of the specimen as shown in
for the control of the camera interface card, the calcula- Fig. 4. Eight is a convenient number (multiple of two) for
tion of the center of each light line, and the transforma- computer operation. To save processing time only eight
tion of the picture message into a digital message, as well horizontal equal-distance lines are scanned in the vertical
as for the selection of the points being scanned. direction. Therefore, there are 64 intersection points on
the specimen panel. This scanning is accomplished by
using the interface card with Assembly language programming.
Theoretical Considerations In Fig. 4 the solid vertical lines (a to h) are the projected
While the 'in-plane' stress distribution plays a key role light lines appearing on the specimen surface. The dashed
in local-buckling strength of plate elements, the in-plane horizontal lines represent the camera scanning lines.
strain is unimportant so far as measuring the local buckling (Actually, they do not appear on the specimen.) The
deformation is concerned. That is, only the out-of-plane camera scans from left to right and from the top down.
deformation is of primary interest. With 256 x 256 resolution scanning from left to right,
Since the video camera can only detect the in-plane the left margin is designated position 0 and the right
movement of the lights, some geometrical considerations margin is designated position 255 (e.g., line 1-1 in Fig. 4).
are necessary. This is illustrated in Fig. 3 where A A is the When vertical light lines intersect with any of these 256
cross section of part of the specimen before deformation, positions, the value in these positions will be '1'. Other-
BB is the cross section of part of the specimen after out- wise, the value would be '0'. The computer then calculates
of-plane deformation, C is the slide projector aimed at the the central position of each light line and stores it in
plate specimen with an angle a , and D is the video RAM. For example, if the first two vertical light lines
camera perpendicular to the surface being measured. (A,B) intersect scanning line 1-1 (Fig. 4), then the left
a a ' is the light line from the projector where a is on A A margin is position '0', and the central positions of light
and a ' is on BB. ab is the line from the camera, b is on lines A and B are 9H and 14H, respectively, where H is

Experimental Mechanics 9 21
508 mm
I
A B C O E F G H
Fig. 4 - - S u r f a c e of the , ~ . . . . I- -I- -I . . . . . . .
specimen being measured. 2 4 - - - I - -I . . . . . . . . .
The vertical bold lines (A-H)
represent the light lines
projected by the slide
projector. The horizontal
dashed lines (1-8) represent
the invisible scanning lines
by the video camera B~ . . . . . Fig. 6(a)--Plate specimen and the measurement area

A B t

080000~--~880000~-~I 0 t = 350 s e c . ~ - /7~-~/


/ , f.----- t =

I I I J
9H 1 4H
i
3

H: BASE 15
Fig. 5 - - E x a m p l e of scanning line. The "O"s are
i4
those positions w i t h o u t light; the " l " s are
those positions with light, totaling 256 !s
positions in each scanning line
i

base 16. This is indicated in Fig. 5. When the specimen


deforms in the out-of-plane direction, the central positions
of those light lines will change. From this change (distance
ab in Fig. 3), the out-of-plane deformation of the specimen
can be calculated.
The test procedures can be summarized as follows. 35 30 25 20 15 I0 5 0
(1) Prepare the required computer hardware and soft-
Displacement (mm)
ware. (2)Place the video camera so that the view line
through the video camera is perpendicular to the specimen. t = 30 sec
(3) Position the slide projector so that the light rays have Fig. 6(b)--Deformation history o f t h e center cross section
an angle c~ of approximately 45 deg (Fig. 3). (4) Through
the monitor, adjust the foci of the video camera and the
projector and center the measuring area. (5) Calibrate the
system. (6)Start the control system and the loading.
Print out the history of the deformation when the test is
completed. specimen is shown in Fig. 6(a). In this test, total time
duration was four minutes. The deformation history of
the center cross section for every 30 seconds is shown
Test Examples of Application
in Fig. 6(b).
A model system including a video camera with 256 x
256 resolution and an Apple II personal computer was
Test 2--Thin-Walled Steel I Beam Subjected to
used in two experiments to demonstrate its applicability.
In these tests eight light lines were projected from a slide
Patch Loading
projector on the plate specimen so that the motion of 64 In this experiment a welded I beam was subjected to
points (8 x 8) are scanned. The maximum scanning speed patch loading. This is shown in Fig. 7(a). The total
is approximately 30 pictures/s. Because the emphasis in duration of this test was 60 seconds. The deformation
this paper is on the approach, detailed test data such as history of the center section for every 10 seconds is shown
the force-deformation curves are not given. in Fig. 7(b). This figure shows clearly that the web of the
beam was crippled near the top (compression) flange at
Test l--Thin Aluminum Plate Under Uniform the early stages of loading and that the final failure was
associated with the second mode.
Compressive Loading
In this experiment the test plate was cut from a sheet of
aluminum plate with dimensions 20 in. x 10 in. x 1/8 in.
Summary and Discussion
Prior to testing the plate was white coated on one surface A simple computer-interfaced optical measuring system
where the light lines were to be projected. The plate was for the dynamic out-of-plane deformation (the local
then placed vertically under a testing machine. The test buckling) of a thin-walled steel structural-plate element

22 9 March 1988
508
was developed. The advantages of the current system
include the ability to measure the dynamic deformation,
the automation of the test process and data acquisition, VVVITI-1
and the low cost. -l- . . . . . . . . 7
The current system, although extremely inexpensive, If Meosurin 9 AreQ
has the following drawbacks for serious local-buckling
study. (1)The system is best suited for those types of (178 mm X 406 mm)
tests where each scan line has similar deformations, i.e.,
the curvature change is only significant in one direction. -Welded I Beom
(2) The area that can be measured is currently limited by
the range of the camera. (3)The accuracy is currently
limited by the degree of resolution of the camera. Fig. 7(a)--Welded I-beam and the measurement area

i
t=0

\ 4
Fig. 7(b)--Deformation history of the center
cross section

6 ~

7 ~

25 20 15 i0 05 0 -5 -i0 -15 .-20 -2

Displacement (mm)

~t = I0 sec

Modifications of the current system are being made as UI: GRATED I,,iODEL
shown in Fig. 8. Major improvements include the use of
two video cameras, a color monitor, and the painting of [8M-PCIX~////I ..... I
light points directly on the surface of the specimen so SPEC IIIEH
that three-dimensional deformation can be traced and
measured. To do this an IBM PC system and two DT2803
image-processing chips are used. The updated system will
be used to investigate the dynamic-local-buckling processes
of thin webs and other plate elements.

Acknowledgment ~ o.MERAS~
This study was supported in part by a grant from the
National Science Foundation (ECE8516471).
Fig. 8 - - T h e updated system

References
1. Novak, P., "'Sterophotographic Study of Impact on Beams and
Frame, "" EXPERIMENTAL MECHANICS, 16 (2), 66-70 (Feb. 1976).
2. Kobayashi, A . S , Manual on Experimental Stress Analysis, 3rd Ed., 4. Ciareia, S., "Building the Micro D-CAM Solid State Video
SESA, 62-69 (1978). Camera, "" Part 2, BYTE, 8 (10), 67-86 (Oct. 1983).
3. Ciarcia, S., "'Building the Micro D-CAM Solid-State Video 5. "Optic R A M IS32 Data Sheet," Micromint, Inc.
Camera," Part 1, BYTE, 8 (9), 20-31 (Sept. 1983). 6. "Micro D-CAM IBM PC Version Users Manual, "' Micromint, Inc.

Experimental M e c h a n i c s 9 23

You might also like