Thermoelastic Forum Vol.1, No.5, October
1994
First Words
R&D Side
Tech Tips
Events
New Products
Tid Bits
The UW Corner
Much has happened since the last issue of the Thermoelastic Forum. Stress
Photonics completed a NASA Phase II SBIR contract that involved the development
of a high-speed Thermoelastic Stress Analysis (TSA) array camera. This prototype
camera was successfully used in a Stress Photonics Air Force Phase I project,
which in turn resulted in the award of an Air Force Phase II SBIR. Under
the Phase II contract, Stress Photonics will be using the array camera to
study high-temperature fracture mechanics of isotropic and anisotropic materials.
(Please refer to the R&D Side column in this issue of the Thermoelastic
Forum for details of this new research.)
Development of the array camera has lead to another bit of recognition.
Each year R&D Magazine honors the 100 most technologically significant
new products of the year. This year, Stress Photonics, together with NASA
Langley, was awarded the "coveted R&D 100 Award" for the DeltaTherm
array camera.
In this issue of the Thermoelastic Forum don't miss (1) new SPATE* and VPI
imaging software, DeltaVision, featured in New Products; (2) the application
of fracture mechanics equations to Thermoelastic Stress Analysis data in
the R&D Side; (3) the University Corner featuring joint research done
by the University of Illinois Champaign/Urbana and Stress Photonics.
If you would like to contribute a comment, an article, or share some TSA
advice, use the response card provided or write to the Thermoelastic Forum
at Stress Photonics.
*SPATE is a trademark of Ometron Ltd. London.
By Jon Lesniak
In the December '92 issue of the Thermoelastic Forum, I described techniques
for measuring stress concentrations in an article entitled "Stress
Intensity Measurement Through Image Deconvolution." That article did
not directly address stress intensity factors by the fracture mechanics
definition. This article, however, offers a method for directly measuring
stress intensity factors. It was found that, unlike the stress concentration
measurements, stress intensity factors can be accurately measured without
image deconvolution. In work supported by an Air Force (WPAFB) Phase I SBIR
grant, a reliable quantitative method for measuring stress intensity factors
from TSA data was developed. Advancements to the works of Peter Stanley
and Janice Dulieu-Smith include
William sought Airy's stress functions under the coordinate system described
in Fig. 1 for a crack in a plate.
By taking the appropriate derivatives of the stress functions and combining
sr and sq the following sum of the principal stress shapes are derived.
Notice, only the l=-1/2 term of Eq. 1a is of significance at the crack
tip where it is singular. For this reason, in classical fracture mechanics,
the other terms are ignored. The shape of the l=-1/2 term is referred to
as the cardioid curve. The stress distribution for many higher order terms
of Eq. 1b - 1d increases as r increases.
The stress intensity factors are derived from the l=-1/2 terms as
However, it is difficult to measure data solely near the crack tip where Eq. 1a is dominant because of
Away from the crack tip, fundamental stress patterns resembling shapes
other than the cardioid curve will be present. Therefore, it is important
for the stress measurement technique to account for all possible fundamental
stress terms.
For the reasons mentioned above, a new technique of determining stress intensity
factors from thermoelastic data was derived. Putting the terms of Eq. 1
in matrix algebra form, the total stress field can be written as a combination
of the fundamental stress shapes,
This can be written more simply as
The least squares algorithm applied to Eq. 4 yields
To simplify the least square algorithm, the basis is orthogonalized and
normalized over the fit window allowing for rapid application and re-application
of the fit. From the solution of Eq. 5, the values of {C} can be determined
and KI is determined from Eq. 2.
The SPATE 9000 camera was used in the stress intensity quantification study.
Simple pin-connected 1018 steel specimens were used in the stress intensity
verification study. Three specimen geometries were used with EDM notches
of a/w ratios of 0.1, 0.3, 0.5. The specimens have a thickness of 1.6 mm
and are considered to be in a plane-stress state.
Figure 2a, showing the fit window, is the close-up TSA image for a/w = 0.1.
Figure 2b shows the out of phase image and the effects of the anomalies.
The data around the crack tip is not used in the fit for reasons previously
described. The fit area is about 10 mm square. In-phase refers to the fact
that the oscillating thermal signal is synchronized with the applied oscillating
load. Out-of-phase signal occurs when inelastic events or heat transfer
is present. From the out-of-phase image the minimum acceptable radius of
the fit window can be determined.
Figure 3 shows the ability of TSA to quantify stress intensity factors.
The quantified values of Y, as defined in Fig. 3 for the three geometries,
are plotted vs. a/w for the four-term and one-term fit estimations and the
theoretical solution.
pdf file for
this issue of the Thermoelastic ForumThe 4-term fit, and a more appropriate use of crack tip data, affords
significant improvement in stress intensity measurements. Simulations indicate
an equal accuracy in determining mixed-mode stress intensity factors.
TSA directly measures the stress intensity factor. It does require knowledge
of the thermoelastic response of the material (i.e., the thermal properties),
but these can be easily determined for each material from a simple uniaxial
bar experiment, the results of which are valid for all applications of that
material.
For more information contact Jon
Lesniak.
By Brad Boyce
Have you ever started a scan and then found out half way through that you
had selected the wrong scan parameters, resulting in an inadequate scan?
If so, here is a suggestion that just might help.
First, do a Point Scan with the same sampling time, time constant, and acquisition
parameters that you intend to use for your frame scan.
A Point Scan provides a line plot of the stress at a point over time. The
Point Scan will quickly and graphically display the signal to noise ratio
for the sampling parameters you have chosen.
For the Point Scan, select a point of nominal stress. (We like to use the
point we used to phase in the lock-in.) Then test that spot with a 100 sample
Point Scan. The resulting plot should have a magnitude of about 500 data
acquisition units (+2048 is positive full scale). If your acquisition parameters
are set too conservatively, your plot will be a nearly perfect straight
horizontal line. If your acquisition parameters are not conservative enough,
the Point Scan data will plot as a widely varying line with a constant offset
from zero.
Soon you will be able to recognize the proper setting of the acquisition
parameters in the familiar 100 sample Point Scan. We recommend that you
always set the 2D-Plot parameters to cover the full data acquisition range
from -2048 to +2048. With these parameters, the plot will always look the
same and you will more quickly get a feel for how the plot should look.
Want more scanning tips? Call, send, FAX or e-mail in your ideas or questions
to Stress Photonics in care of the Thermoelastic Forum.
The Spring SEM Conference was held this past June in Baltimore, Maryland.
On the trade-show floor Stress Photonics displayed an array-based TSA camera.
Two papers were presented at the conference that dealt with thermoelasticity.
"A High-Speed Differential Thermo-graphic Camera" by Lesniak
and Boyce
This paper describes the development of an array-based TSA camera with short
image acquisition times (10 to 15 seconds). The signal to noise ratio of
staring-arrays and scanning-detectors was compared with excellent results.
"The Thermoelastic Response of a Thin-Walled Orthotropic Cylinder
Loaded in Torsion" by Dulieu-Smith and Stanley
In this paper, TSA was performed on a thin-walled wound aramid-epoxy cylinder.
The next SEM conference, "Integration of FEA and Structural Testing with Rapid Prototyping," is to be held November 7-8 with an optional tour November 9th, 1994 at the Hyatt Regency in Dearborn, Michigan.
DeltaVision TM software by Stress Photonics allows SPATE, VPI and DeltaTherm
1000 users to take advantage of powerful UNIX workstation computers, software
and network resources to process and present data quickly and easily.
Moving image processing to the UNIX workstation allows the user to take
advantage of valuable network resources, such as
DeltaVision software is written in Visual Numerics' popular PV-WAVE Command
Language making it portable between platforms and easily customized and
adapted to work with other popular engineering workstation software.
DeltaVision was written specifically to simplify data visualization.
DeltaVision features
DeltaVision's graphical user interface conforms to the Motif and X-Windows standards making it easy to learn and operate. The windowing environment
The software offers options to output or "save" graphics to several popular file formats including tiff, PostScript, pict and PCL. These files can be printed directly or easily transferred to an IBM or Macintosh format for word processing. DeltaVision images have already been used for journal articles, reports and newsletters such as this.
Hot Detectors from Stress Photonics
With Hot Detectors from Stress Photonics you can now dramatically improve
the scanning performance of your SPATE. Stress Photonics has just upgraded
Pratt & Whitney East Hartford's SPATE 9000 head with the latest in high
performance infrared detectors. With the new detector installed, the infrared
signal is 12 times stronger! This new improvement assures 4 times better
or 12 times faster scans. For detailed information please refer to the insert
in this edition of the Thermoelastic Forum.
If this improvement might help your productivity, simply contact Stress
Photonics or Ometron, or send in the reader response card, and we'll tell
you more about Hot Detectors.
By Dr. T.J. Mackin Department of Mechanical and Industrial Engineering,
The University of Illinois, Urbana, IL
Composite aircraft wing-skins are manufactured to a prescribed thickness
using a lamination process, wherein fabrics of fiber reinforced material
are stacked atop each other, followed by matrix infiltration. The interface
between layers is matrix rich and especially susceptible to delamination
damage. This is a particularly vexing problem in the aerospace industry,
since sub-surface delaminations are not easily detected.
Modern aircraft are designed to carry substantial loads in the wing-skins:
tension on the lower side, compression on the upper side. Impact events
generate stress waves that, in effect, probe the interface region of a laminated
structure. Thus, impact, such as tools dropped on the wing during routine
service, can generate insidious, sub-surface interface delaminations. Delamination
damage greatly reduces the compressive strength of a laminate and must be
avoided or contained in order for the wing to carry its design load. Modern
composite materials employ stitching to improve interface toughness. The
stitching holds the layers together such that a delamination is halted at
the stitching line.
A controlled study of impact damage was conducted to evaluate the retained
compressive strength of carbon fiber reinforced epoxy laminates. Plates
of composite wing-skin were subjected to a range of impact conditions, spanning
changes in impact velocity, energy, and radius of the striking tip. Certain
impact conditions resulted in sub-surface delaminations with little or no
trace of damage at the impact location. After impact, thermoelastic stress
analysis was used to evaluate the damage. The specimens were loaded under
cyclic compression, at a maximum load equal to 10% of their compressive
strength. Thermoelastic images revealed a region of damage coincident with
the location of impact, Figure 1, where the extent of damage was related
to the striker geometry and impact energy. Notice the visible stitch lines
appearing horizontally across the image.
pdf file for
this issue of the Thermoelastic ForumThermoelastic analysis was also used to evaluate stress redistribution
across the specimen. Using this information it is possible to quantify the
damage and relate post-impact compression strength to the extent of the
damaged zone. It is postulated that, during post-impact compression testing,
the damaged zone suffers from an early onset of compressive buckling, which
cascades into the remainder of the specimen.
For more information contact Dr. Mackin at the University of Illinois.
TSA image collected at the University of Wisconsin-Madison by Jon Lesniak
of Stress Photonics.
