Special Report 88-9 US Army Corps

July 1988 of EngineersCold Regions Research &Engineering Laboratory

Behavior of materialsat cold regions temperaturesPart 1: Program rationale and test plan

Piyush K. Dutta

DTIC

SEP6lg 8 -

Prepared tor88 0OFFICE OF THE CHIEF OF ENGINEERS 8 10 6Approved for public relcase; distribution is unlimited.

'IIUNCLASSIFIED

SECURITY CLASSIFICATION OF THIS PAGE

Form ApprovedREPORT DOCUMENTATION "'AGE OMB No 0704 0788

_ Exp Date Jun30 1986l 1a REPORT SECURITY CLASSIFICATION lb RESTRICTIVE MARKINGSUnclassified

2a SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/AVAILABILITY OF REPORTApproved for public release;

2b DECLASSIFICATION/DOWNGRAL)ING SCHEDULE distribution is unlimited.

4 PERFORMING ORGANIZATION REPORT NUMBER(S) S MONITORING ORGANIZATION REPORT NUMBER(S)

Special Report 88-9

6a NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a NAME OF MONITORING ORGANIZATIONU.S. Army Cold Regions Research (If applicable)and Engineering Laboratory CECRL Office of the Chief of Engineers

6c. ADDRESS (City, State, and ZIP Code) 7b ADDRESS (City, State, and ZIP Code)

Hanover, New Hampshire 03755-1290 Washington, D.C. 20314-1000

Ba. NAME OF FUNDING/SPONSORING 8b OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION (If applicable)

%. 8c. ADDRESS (City, State, and ZIP Code) 10 SOURCE OF FUNDING NUMBERS

PROGRAM PROJET TASK WORK UNITELEMENT NO. NO 4A7627 NO ACCESSION NO

6.27.30A 30AT42 SS 019

11 TITLE (Include Security Classification)'-.. Behavior of Materials at Cold Regions Temperatures

Part 1: Program Rationale and Test Plan12 PERSONAL AUTHOR(S)E Piyush K. Dutta13a TYPE OF REPORT 113b TIME COVERED 14 DATE OF REPORT (Year, Month, Day) 15 PAGE COUNT

FROM TO July 1988 7216 SUPPLEMENTARY NOTATION

17 COSATI CODES 18. 4ACT TERMS (Continue on reverse if necessary and identify by block number)

F;ELD GROUP SUB-GROUP Cold regions Low temperaturesCompression~tension testing Materials testing

1e19 ABSTRACT (Continue on reverse if necessary and identify by block number)

Newer materials and products are being constantly added to the Army's inventory. Cold regions climatic• conditions should not impair the reliability and durability of these new systems. This report discusses

the rationale of the test program being undertaken at CRREL to evaluate material behavior at lowtemperatures.

20 DISTRIBUTION /AVAILABILITY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION

" UNCLASSIFIED/UNLIMITED El SAME AS RPT El DTIC USERS Unclassified22a NAME OF RESPONSBLE IND:VIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL

Pivush K. Dutta 603-646-4100 CECRL-EA

DD FORM 1473, 84 MAR 83 APR edition may be used until exhausted SECURITY CLASSIFICATION OF THIS PAGEAll other erjiton ArP o UNCLASSIFIED

N..........................................................."- ",. .",",..-.... ,...

PREFACE

This report was prepared by Dr. Piyush K Dutta, Materials Research Engineer,of the Applied Research Branch, Experimental Engineering Division,U.S. ArmyCold Regions Research and Engineering Laboratory. Funding was provided byDA Project 4A762730AT42, Design, Construction and Operations Technology forCold Regions, Task SS, Service Support, Work Unit 019, Behavior of Materials at

-' "Low Temperatures.The author thanks Dr. Ronald Liston of CRREL and Dr. Harold Lord of Michi-

gan Technological University for technically reviewing this report.The contents of this report are not to be used for advertising or promotional pur-

poses. Citation of brand names does not constitute an official endorsement or ap-proval of the use of such commercial products.

., .- *,.

- ..

S" %-

S

0?r

"" ii

0".

.. . . .- . . p F . . - ~ .".-.-

CONTENTSPage

A bstract ................................................................................. iP reface ................................................................................. iiIntroduction ............................................................................ 1

*'. Low temperature properties of manufactured materials-background ........ 2Metals .............. ................... ................... 2Plastics/polym ers ................................................................. 5C om posite ........................................................................... 6

Fracture property tests for low temperature application of materials ........... 8- Early fracture test approaches .................................... 8

Fracture m echanics test ........................................................... 11Fracture toughness tests for polymers and composites ................ 16Influence of cyclic load on fracture toughness ................................. 16Strain rate effect on fracture ...................................................... 17

T est plan ............................................................................... 19R ationale ........................................................................... 19Experim ental program ............................................................ 19Expected information ......................................... 25

* Selected bibliography ................................................................. 26Appendix A: Low temperature behavior of some structural metals

and alloys ........................................................................ 31Appendix B: Low temperature behavior of some selected polymers ............. 45

ILLUSTRATIONS

Figure1. Photograph of T-2 tanker that failed at pier ........................ 22. Impact test results for a steel pipe material .................................. 33. Typical stress-strain curve of a body centered cubic class metal at

decreasing temperatures ............................................... 34. Simplified deformation behavior maps of FCC and BCC material ....... 45. Energy-temperature curves obtained by Charpy V-notch tests of

twelve steels .......................................... 46. Therinomechanical curve of a simple polymer ............................ 57. Schematic presentation of viscoelastic properties of polymers.......... 58. Stress-strain behavior of epoxy resins at different strain rates, i and

* tem peratures, T .......................................................... 69. Simple classification of composite system .................................. 7

10. Orientation dependence of composite property .............................. 811. Temperature dependence of composite property ............................ 812. Temperature effect on the fracture stress .................................... 913. Temperature effect on ductility ............................................... 914. Simple beam Charpy impact machine ....................................... 1015. Centrally cracked panel under uniform axial stress ...................... 1116. Plain strain and plane-stress fractures ..................................... 1317. Typical load-crack displacement curve ......................... 1518. Transition of fracture toughness ............................................. 1519. Test program schematic ............................. ....... 20

* 20. Device for quasi-static slow and fast loading ............................... 2021. Split Hopkinson pressure bar applied to fracture dynamics .............. 22

.. iii

', TABLES

Page

Table1. Mechanical equipment and instrumentation requirements for low

temperature test of materials ........................................... 24

40

7iv

%

".J.

.. iv

4 Behavior of Materials at Cold Regions Temperatures

Part 1: Program Rationale and Test Plan

FIYUSH K DUTTA

INTRODUCTION

Cold regions engineering has a general need for the evaluation of materials atlow temperatures. Performance of many structures and components is seriouslyaffected when the weather becomes very cold. At low temperatures materials tend tobecome hard and brittle; as a result, legitimate concerns are raised about theirreliability and safety in such cold weather. The disastrous accident of space shuttleChallenger because of low-temperature-induced failure of O-ring seals is a grimreminder of such problems. Moreover, newer and lighter materials and products

"-" .for engineering structures and systems are rapidly replacing the older and heav-ier materials and components. There is a growing need to evaluate these materialsat low temperatures. In particular, the U.S. Army needs to evaluate new metal-substitute polymers and polymeric composites in cold climates for its new High-Technology Light Division.

Because of these needs, a low temperature (down to -60°C) material testing pro-gram is being esttblished at CRREL. The testing program includes a variety oftests ranging from ordinary compression-tension testing of specimens to loadingat very high 4train rates in a Hopkinson split-bar system. Destructive, non-destructive and fracture mechanics tests are also planned to be performed at lowtemperatures. This report discusses the rationale for this program and outlines thetest plan.

Cold regions material problems are many. Brittle-fractures initiated by lowtemperature or high strain rate (or both) have at times caused large-scale damage.In cold weather (00 to 50C) merchant vessels have broken in two while in harbor(Fig. 1), bridges have collapsed, and pipelines and gas storage tanks have rippedopen (Parker 1957). Exploration and production of oil and gas have begun to takeplace in colder climates (down to -60C), and in deeper and rougher water; these

rigorous conditions aggravate the brittle fractu-e hazard. At present, the knowl-edge necessary to prevent brittle fracture is far from complete.

% Polymeric composites also have the potential for catastrophic failure at cold re-gions temperatures. These materials tend to develop brittle failure characteristics,depending on their composition and manufacturing processes. The development ofbrittle failure in these materials accelerates with lower temperature, material de-fects (notches or cracks), fatigue, and high-strain-rate loading.

In steel, a small reduction within a narrow temperature range called the transi-tion temperature can cause a sharp decrease in fracture toughness. A preexisting

* crack that has become arrested at normal temperature and is thus tolerated can be-come self-propagating at low temperature and may cause catastrophic structuralfailure.

V, CRREL's low temperature material testing program will perform both quasi-static and high-strain-rate loading fracture toughness tests on selected materials.Tests will be done on a variety of materials, including plastics, composites, andmetals. Some materials will be subjected to cyclic loading at lower temperatures

0Z

;r,; .

0 .. On

Figure 1. Photograph of a T-2 tanker that failed at pier.

before the fracture mechanics tests are performed in order to evaluate propertydegradation (fatigue) in service.

Quasi-static tests will be performed using the MTS and INSTRON materialtesting machines now available at CRREL. A high-strain-rate loading facility us-

,W ing the Hopkinson split-bar system has been installed and suitably instrumentedto record and analyze the high-speed stress waveform data.

In the following sections the general background of low temperature materialbehavior and critical physical and engineering property data are reviewed anddiscussed. An outline of the proposed program for investigating the critical phys-ical properties of a few selected materials at low temperature is also presented.

LOW TEMPERATURE PROPERTIES OFMANUFACTURED MATERLIS-BACKGROUND

Most manufactured materials exposed to low temperatures show a substantialloss of useful structural properties. Generally speaking, as the temperature is low-ered, the hardness, yield strength and modulus of elasticity increase, but fracturetoughness/impact strength, fatigue strength, Poisson's ratio, thermal expansioncoefficient and specific heat decrease.

Metals* At low temperature the most significant property change in metals is the in-

crease in brittleness. A metallic object or structure will shatter or fracture when* subjected to stresses (especially from impact) that are allowable at normal temper-

atures (Fig. 2). The metals that have face-centered-cubic lattice structure, e.g.nickel, copper, aluminum, lead and silver, show some ductility at low temperature.But a larger group of body-centered-cubic class metals, e.g. iron, chromium, mo-

0.: lybdenum and tungsten, show a marked decrease in ductility.

2

.,*0'

4r1

% I" 30 -

-BO -60 -40 -20 0 20 40 60 80Rooi temperature Extreme cold Temperalure *F)

Figure 2. Impact test results for a steel pipe material.

200'F

0 -150

V)

Failure

-100

Figure 3. Typical stress-strain70 curve of a body centered cubic

class metal at decreasing tem-Strain - peratures.

Loss of ductility in a metal can be observed by examining its low temperaturestress-strain relationship. As the temperature is lowered, both the yield point(where ductility begins) and the ultimate strength point (where failure occurs) mayshift to a higher stress value, but fracture may begin at a much lower strain value(Fig. 3).

* Figure 4 shows the general aspects of temperature-dependent mechanical behav-ior of both face-centered-cubic (FCC) and body-centered-cubic (BCC) metals. Notethat at the lowest stresses a specimen deforms elastically; increasing stress even-

V" tually brings the specimen to the limit of the elastic region where plastic strain(permanent deformation) or brittle fracture occurs. Figure 4a shows the behavior ofFCC materials. Note that the yield strength increases with decreasing temperature,

• but the increase in ultimate strength is larger at lower temperatures. Figure 4b il-lustrates the BCC behavior. This figure shows a brittle region where the specimen

Z' fails before any significant plastic deformation occurs. This brittle region is ex-tended to higher temperatures by several factors, namely (1) sharp cracks or otherstress concentrators, (2) higher strain rate, and (3) impurities, e.g. higher carbon

0 contents in steel.

3

1040 - 1 4 1

20-

,, ' 10-3 Ductile I CAC

80-

.,": C 6W D

(1) Elastic Strain 40-

I __ ___ _ 20--

o Br~ftle [a'lure- -120 -80 -40 0 40 80 120 160 200

,1" 1 Ductile -- Testing Temperoture (F)

' Composition

10 Steel C Mn Si P S Ni Cr

0 (2) A 0.17 0.53 0.25 0.011 0.02 3.39 0.06(I B 0.18 0.82 0.15 0.012 0.03 0.04 0.03Q2

C 0.18 0.73 0.07 0.008 0.03 0.05 0.03D 0.20 0.33 0.01 0.013 0.02 0.15 0.09

TITm

, Normolized Temperature

Figure 4. Simplified deformation be- Figure 5. Energy-temperature curves ob-havior maps of (1) FCC material and tained by Charpy V-notch tests of steels.(2) BCC-material.

The fracture mode transition from relatively ductile to relatively brittle is be-lieved to occur because of the very rapid rise in viscosity with decreasing tempera-ture. In most materials the flow stress is low at high temperatures. The stress con-centrations around a dislocation can always be relieved by local plastic deforma-tion and there is no brittle fracture. At low temperatures, the yield stress for slip ishigh so that deformation will start by microcrack production and subsequent brittlefracture (McChntock and Ali 1966).

Because of the onset of brittle fracture, most standard constructional carbon* steels cannot be depended upon at the temperatures encountered in cold regions.

Figure 5 shows energy-temperature curves obtained by Charpy V-notch tests of afew steels having different chemical compositions. The curves clearly show thetemperatures where transition in the fracture mode occurs (McClintock and Ali1966). However, continued research in steel development and processing has re-sulted in a new class of high strength, low alloy steels that can be used in low tem-perature environments (Baldy 1976).

The low temperature properties of a number of metals have been summarized in% Appendix A. These properties indicate that stainless steel is probably the best-suited

ferrous metal for cold use. Unlike other steels, stainless steel has no transitionfrom tough to brittle in a cold environment. The properties of aluminum do not de-

0 grade at low temperature, and it is a preferred material for many cold weather ap-

4

.0,

plications. The ductility of cold-worked copper actually increases at low tempera-ture, which makes it a very good metal for use in the cold regions.

Plastics/polymersPlastic materials generally become more brittle at low temperatures, but they are

not as consistent as metals. The tensile strength of plastics increases very mark-edly at lower temperatures. For example, a nylon material increases in tensilestrength from 7400 psi at 21'C to 13,000 psi at -57 0C, but its impact strength decreasesfrom 16 if-lb at 21 0C to 0.9 ft-lb at -40'C (Hansen 1960). Some plastics, for examplepolyethylene, a thermoplastic polymer, remain tough at temperatures as low as-73°C. However, serviceability of rubber components, e.g. tires, inner tubes, cable,hose, bushings and seals, is seriously affected by low temperature. Rubber developsbrittleness and loses flexibility; its loss of resilience is associated with changes inhardness, volume and coefficient of thermal expansion.

Both time and temperature influence the behavior of polymers. Figure 6 showsthe typical thermomechanical behavior of a simple polymer. At low temperature itbehaves as a glass. As the temperature is raised, it becomes less brittle. The upperlimit of the glassy region is called the glass transition temperature, Tg. Com-monly observed values of Tg are in the range of-53C to -97 0 C (Read 1983). In theglassy region, polymers typically have a modulus of elasticity of 105 to 106 psi if or-ganic and up to 107 psi if inorganic. Poisson's ratios are 0.25 to 0.4. In the rubberyregion the modulus drops to 102 to 103 psi (McClintock 1966).

The stress-strain behavior of polymers can be represented by a system of dash-pots and springs as shown in Figure 7. At low temperatures the dashpots becomemore and more frozen, resulting in very little flow. Near the glass transition tem-perature the flow is established and the material behaves viscoelastically (Fig. 8).During viscoelastic deformation, the stress-strain relation is strongly dependenton the strain rate . A high strain rate leaves the polymer little time for plastic flow.The higher the strain rate, the more elastic and stiff a material becomes. At lowerstrain rates dissipative effects connected with viscous flow result in hysteresis of

EL ASTIC

F, IT) ~I

Rubbery/

Flow /

Leathery Rubbery I Liquid VISCO-// I/ELASTIC

F2I)e (MEMORY)II F2 (TI (} ~-q , IT I

Typicol Amorphous Polymer

I1, T I

200 400 600Temperature (K) a

Figure 7. Schematic presenta-Figure 6. Thermomechanical curve of a tion of viscoelastic propertiessimple polymer. of polymers.

5

..................

T-4 2 T77K~~~.~ ~ '0 2017

Sec Sec

-/ rfrrcture

T-293Ki-0 36 -Sec

N ' / .

:44

x .'" g . O 6 O 10 12 %

S7RAIN E

Figure 8. Stress-strain behavior of epoxy resins at differ-. ent strain rates, i, and temperatures, T (after Hartwig

1979).

the loading and unloading curves. Figure 8 also shows that at low temperaturesepoxy resins exhibit a linear elastic behavior up to the fracture point; there is nohysteresis effect and no strain rate dependence (Hartwig 1979).

The behavior of polymers is exceedingly complex. The molecular structures thatinfluence this behavior are (1) branchinig, (2) chain length (molecular weight), and(3) crystallinity. The influence of low temperature on these parameters and the re-sulting change in behavior have been studied extensively in recent years (Hartwig1979, Frank 1979). Appendix B ,ummarizes the low temperature behavior of anumber of polymers.

, ComTositesThe term "composite" refers to materials having overall properties that are some

average of the properties of several distinct components. One of these components iscontiguous and forms a matrix interfacing with the reinforcing elements. In lowtemperature design, composites are particularly attractive because of their low ratio

0' of thermal conductivity to elastic moduluz; oi strength and their high ratio of elasticmodulus to density and strength to density.

Theoretically, both filamentary-reinforced materials, such as fiberglass-epoxy,and aggregates, such as concrete, fall within this definition. Figure 9 gives a sim-ple classification of composite systems. Discussion in this section will primarilyrelate to polymeric composites.

*The material properties of fiber-reinforced composites depend on (1) matrix type,(2) fiber type, (3) fiber fraction, (4) fiber orientation, (5) fiber distribution, (6) fiber

%

.0I6'

"Q . "... . -,'....'.-, .. ", . .. "..-". .. '. .- . '. .. .. .'.".., " . . .... -'.2 -. , -•. .. --: -'-2 .'.. ': .- ,,.,'.',: ' ,,.6,'

V~~~~~~ ~ ~ ~ ~ v~w w -J-r~ W- - 0XrawJ J -J . Qw-p7- '

iCOMPOSITES

reinforced reinforced inforced

products -e.g. rod, fabrication aspipes, angles, etc. per component

nee mareinorceng

components Low-modulus High-modulusreinforement, reinforcement,e.g. fibers of advanced composites

glass, cotton, etc. e.g. fibers of

L'" This produt will bdesig, need mchininga

Figure 9. Simple classification of composite system.

* strength and (7) temperature. Most of the properties are anisotropic and follow thegeneral trend of increase in elastic modulus, tensile strength, flexural strength,and compressive strength with decrease in temperature. The properties of compos-ites can be controlled by construction techniques, and customized material devel-opment is possible. Because composites are relatively new structural materials,their use was initially restricted to selective replacement of specific metal compo-nents, but this conservative approach is gradually being abandoned.

Low temperature (to the cryogenic level) behavior of composites has been exten-sively reviewed by Kasen (1975, 1983). He has observed that there is no systematicdata base for lower temperatures. Existing data show extreme variability instrength properties as a result of embrittlement of matrix resins. Efforts to developmatrices with improved low temperature toughness continue; however, achievingthis objective remains elusive (Hartwig and Evans 1982). The majority of the cur-rent data on polymeric composite strength are for room temperature and for theboiling points of cryogens.

The large variability in composite strength data is due to embrittling of the poly-mer matrix in cooling. Variability is also a strong function of the quality of lami-

nate manufacture, for example, percentage of void- and resin-rich areas. Qualitycontrol during manufacture can therefore largely control the strength properties ofcomposites (Kasen 1983).

The anisotropy of the reinforcement results in anisotropy in the properties of thecomposite (Fig. 10). Therefore, composite data are customarily related to fiber ori-entation. Figure 11 shows the temperature dependence of two glass-fabric-rein-forced epoxy composites in two different orientations (Ledbetter 1979).

7

nat S. -k ., manufcture fo ea e are g P A a A 1ares Qait

IL.%- - NK -MA U .

,% Fu 1.rntnen 6- -G I1CR

-- ! CR

I Normal

14 Direrlon

.'.

0 0 00 200 300FILL. GPa TEMPERATURE. K

Figure 10. Orientation depend- Figure 11. Temperature de-ence of composite property. pendence of composite pro-

perty.

FRACTURE PROPERTY TESTS FOR-e LOW TEMPERATURE APPLICATION OF MATERIALS

Because of economic competition, weight premiums, and the availability ofmaterials with high yield strength, structural engineering designs are tending toincorporate higher nominal stresses. However, this trend has also increased theprobability of catastrophic failure. Safe operation of structures at high stresses re-quires an accurate definition of the fracture resistance of materials to ensure thatan adequate safety margin is maintained in material performance to balance theeffects of design refinements, fabrication, quality and reliability of inspection.

Fracture resistance depends highly on the conditions related to environment,constraint and strain rate. Fractures can occur at stress levels below yield stress orabove yield stress, depending on flaw size, section thickness and strain rate.

% In order to discuss the effect of temperature on fracture let us recall Figure 4b andshow it schematically in Figure 12. It will be seen from Figure 12 that a material,when stressed at temperature Ti, will deform first by plastic yielding. If it were tostrain-harden to the curve ao , it would then undergo brittle failure at stress crF. Ifthe temperature Ti is below TA, the nil ductility temperature (NDT) he materialwill simply fail without any plastic strain when the stress reaches aF. We havealready mentioned that BCC materials are far more sensitive to temperature thanFCC materials. Strain rate tends to increase the yield stress ay , displacing the Gycurve upward as shown by the dashed line in Figure 12. This will lead to brittle

* failure at higher temperature since TA will have increased.

Early fracture test approachesThe foregoing discussion shows that the fracture properties of a material depend

on test conditions (temperature, strain rate) and specimen conditions (flaws).Early test methods like Charpy impact testing (ASTM 1966, Standard E-23) were

_ therefore designed to introduce a given flaw into a specimen and to test that speci-

8

%.%

'DT

. Strain-irate

I I'aY" e irate

r ef

Str inn

T2

e m p e ra t u r e - fracture stress.

2 2

StrainI T2

Temperature

Figure 13. Temperature effect on ductility.

men at a high strain rate. Both the flaw and the high strain rate increase the nilductility temperature, and the test therefore gives an NDT that tends to be con-servative. The energy under the stress-strain curve is used as a measure that canbe related to ductility. A low energy is indicative of low ductility, as shown inFigure 13a. The energy measured is really the amount of energy used in causing

* plastic flow. Figure 13b shows the difference of energy at two different temperaturelevels, T1 and T2 (T1 < T2 ). The Charpy impact test is a simple and inexpensive testthat is easy to carry out in acceptance testing at the steel works.

The most common specimens employed with the Charpy test in recent years havebeen of the V-notch or keyhole type. A specimen is machined from the plate or otherpiece of material being tested, cooled to the desired temperature, and placed in the

• testing machine in such a way that it is supported on both ends as a beam (Fig. 14).A single blow is applied to the middle of the specimen, opposite the notch; as the im-pactor swings through, the specimen breaks and a recording dial indicates the ab-sorbed energy. Normally, three specimens are tested at a given temperature andadditional tests are made at selected temperatures. A curve showing the mean val-

* ues of absorbed energy versus temperature is plotted as shown in Figure 13b.

9

04 % % %

, % ", , % , - %,% - -,,. , . . % , , % ",% - .. "% -. .. 7 1 " .% 2. , ,. L, _% L. ,- ,, % '

v~r ~ ~ V~ U W % ~.U. W~I WW U ~ WVU.W W w WWW T -Z. ' f KX~ ~~~~ ~'~ YW'W

P1,'of I'MPOC

''J

Fiur 14.. Sipeba Ca m

-MM/

" Figure 14. Simple beam Charpy im-

pact machine.

The Charpy test is used primarily to establish acceptance specifications based onfixed energy levels (say 15 ft-lb of energy at 25°F). It usually shows wide scatterwhen specimens are taken from different areas of the same stock, and there are

MD great variations in the transition ranges. Also, the test has not been very conduciveto formulating stress analysis theories (Witzel and Adsit 1969, Harsem 1969, Hall1969). However, to date it has remained one of the most common methodS of tests toevaluate impact resistance properties of materials.

Another test that has increased in usage is the Naval Research Laboratory(NRL) drop weight test (Pellini and Puzak 1963). This test can be carried out withcomparatively simple equipment at various temperatures and produces very littlescattering in the results (Harsem 1969).

. The drop weight test employs special beam specimens in which a crack is createdin the tensile surface. The test specimen is generally about 14 in. long, 3.5 in. wide,and between 0.5 and 1 in. thick. A weld deposit is placed at the center of the speci-men and notched to provide the initiation source. The test is conducted by subject-

* ing a series of specimens of a given material to a single impact load at selectedtemperatures to determine the maximum temperature at which there is a break (go)or no-break (no-go) condition. The impact load is provided by a guided free-fallingweight with an energy between 250 and 1200 ft-lb, depending on the yield strength ofthe steel to be tested and the size of the specimen. The specimens are not allowed todeflect more than a few tenths of an inch before they reach a stop.

* ASTM (1966) specification E208 describes in detail the provisions for conductingand interpreting these tests. The nil ductility transition (NDT) temperature is de-termined as the maximum temperature at which a standard drop weight specimenbreaks when tested in accordance with the provisions of the drop weight test.

Around 1969, the Naval Research Laboratory developed what was called the dy-namic tear (DT) method of testing with the aim of overcoming the limitations of the

I10

r% %

Charpy V(Cv) test and providing information of engineering design significancein the form of a "stress intensity factor" (Lange and Loss 1969). The test methodwas considered successful in that for very brittle material, where linear elasticfracture mechanics is applicable, direct correlation has been established betweenNDT energy and static stress intensity factors, KIc (discussed later).

Many other test methods have been used in practice and research during the lasttwo decades. A review of all these tests can be found in Hall's (1969) paper on theevaluation of fracture tests.

Fracture mechanics tests.In recent years tremendous interest has been generated in fracture toughness

testing based on linear elastic fracture mechanics. The theory of fracture tough-ness is based on the concepts of Griffith (1920) concerning fracture in brittle mate-rial. Because of much interest and extensive research, the theory of fracture tough-ness will be summarized here to clarify later discussion.

Theory of fracture toughnessAccording to Griffith's theory, when a centrally cracked panel (Fig. 15) is sub-

jected to a uniform axial stress a (psi), the elastic energy loss U (in.-lb) of the platedue to the growth of a crack of 2a (in.) length is given (Nadai 1950) by

U= 7Ca2 a 2 tiE (1)

where t is the thickness of the plate (in.) and E Young's modulus (psi). IfS is thesurface energy (in.-lb) of the two new surfaces created by the crack, then

=4 S =4yat (2)

where t is the surface tension of the material (in.-lb/in.2 ).

-0St

,

2'

- -- Figure 15. Centrally cracked panelunder uniform axial stress a.

%1

6W

Equation 1 shows that if a is increased, more stored elastic strain energy will beavailable to propagate the crack. Obviously, this crack-driving force is opposed byan equilibrium resistance of the material. When the rate of ia-rease of the crack-driving force with respect to crack length equals the rate of increase of the resis-tance, the crack becomes self-propagating. This type of critical condition occurswithout additional external work only when the elastic strain energy released by aminute extension of crack length aa is sufficient to develop a new & of cracklength. This unstable or critical condition can be defined as

a(U-S) - 0

c3a

and the stress at which this unstable condition develops is called the critical stress13c.

'-V By use of the expressions for U and S from eq 1 and 2 and differentiating, thecritical stress ac is given by

''' 1/2c= L (3)

ira

* Although the above argument is true for brittle materials like glass or ceramics,the case of metals is different. For metals a much greater energy is required toplastically deform the grains surrounding the crack tip. However, if a plasticstrain energy term P is added to the expression, the new critical stress for metals

1 can be represented (Orowan 1955, Zinkham et al. 1969) by

ac = [E 2+P)/Ia/ (4)

For metals, the value of P is normally several orders of magnitude larger than thesurface tension of the cracks, so that y can be disregarded.

Irwin (1958) used linear elastic stress analysis theory to show that in an infiniteplate the strain-energy release rate Gc , equivalent to P in eq 4, is related to thecritical stress intensity Kc at the crack tip. Thus disregarding y and substitutingGc for P in eq 4, we have

2 = 2 (5)

where Gc = critical strain energy release rate (in.-lb/in.)Kc = critical stress intensity factor (psi/in.)

CFC = cross-section stress at crack onset (psi)ac = half crack length at crack onset (in.).

Equation 5 is true for an infinite plate. For a plate with finite width W the ex-". pression for the correction factor a (Irwin 1958) is

a [(W/x a4 tan (nac/W)] (6)

12

0%

According to Irwin (1958), eq 5 for a finite-width panel can be written as4

EG C =K 2 = rat21 2 (7)

Since from eq 6

=- tan '-

eq 7 can be simplified to give

EG c =KC2 = W tan- . (8)

The ac term, a critical crack length, generally includes a correction for the sizeof the plastic zone riy at the tip of the crack, i.e.

ac = a c + rj = ac+ K 2 (9)2w2,

fwhere a'c is physical size of crack and ay. tensile yield strength of material.

Note that if a calibration is made of the specimen's compliance with cracklength, the plastic zone is reflected in the compliance measurement in an actualtest, and therefore it is taken into account. If a calibration is not made, the value ofry is estimated through an iterative process by first assuming ac = a'c. and calcu-lating the first Kc value from eq 8.

" Irwin (1958) has shown that Gc or Kc decreases if thickness increases because oftransition from a plane stress to a plane strain condition (Fig. 16). In thin speci-

FatigueCrock

100% Shear

ThinningIn Thickness

Direction

Piane Stress Condition

FatiqueCrock

No Thinning

SN nFigure 16. Plane strain and plane stressPione Strain Condition fractures.

13

!:

" mens plane stress situations are associated with a fracture surface which ap-proaches almost 100% shear. Thus stresses are not developed in the thickness di-rection to resist deformation. The stress intensity factor is associated with planestress or a mixed mode (normal and shear stress). The material property respon-sible for developing resistance to deformation is not well defined in this case.

With an increase in thickness the plane strain condition dominates, and thefracture surface becomes flat, with little or no shear lip. The stress in the thicknessdirection is developed depending on its resistance to deformation.

Now, remembering from eq 7 that for plane stress EG, = K, for plane strain wecan write

. .EG, = Kc (10)

where 4± is Poisson's ratio.We now need to discuss the distinction between KI and K 0 • K is termed the gen-

eral stress concentration factor, and K is a mathematical quantity related to thestress intensity factor of the first mode (the opening mode, with no consideration ofshear but only of cleavage, i.e. a tensile-stress-related fracture opening). Theability of the material to resist deformation in the cleavage or first mode in thepresence of the crack is manifested in the fracture toughness parameter KI, or Gjc

* which may be termed a material constant. The important distinction between KIand Kc is similar to the distinction between stress and strength. Several defini-tions were proposed for Kjc (Strawley and Brown 1965, Strawley 1969). The latestoperational definition by Strawley (1969) is somewhat arbitrary. This defines KIcas the stress intensity at which the crack reaches an effective length 2% greaterthan at the beginning of the test. This is analogous to definition of yield strength interms of a specified amount of plastic strain. Defining K in this manner, we canrewrite eq 10 asa.

. " EG I, = K,2 (1 - p)(

'S.. In the plane strain (thick specimen) test condition there are two general pointsthat are relevant in calculating GIc or Kc (Strawley and Brown 1965). First, it is

t., assumed that no stable crack occurs and the initial crack length is used in thecalculation; and second, the plane strain plastic zone correction term is taken to beone-third of the plane stress term, that is, rly = riy/3 in eq 9, so that

02,;'ac =ac + K C

'6,,

* Equation 8 (after Zinkham and Dedrick 1969) can now be completely modified toshow that

EGj:cKI1- 2)= 'c(1- 2)W tan -- aC +6 (12)

6WW

0~

14

0%-.-.-

-'a. ,.. ., ' ' ' " ' ',-., , ,: .---":-- .--.:.,--:".:..',<, ,-" -.-"., ,Z. .. .. ..,"

1"W WW W V U 'P

where a, = half the original crack ltngth (in.)k= gross stress at the initiation of slow crack growth (psi)

= Poisson's ratio.

In uniaxial tensile tests of an edge-notched or centrally notched specimen, aphenomenon called "pop-in" (a rapid decrease of load and then increase) has beenobserved. Boyle et al. (1962) have found that the "pop-in" load value coincides withthe plane strain parameter Ktc. Figure 17 shows a typical load-crack displacementcurve indicating the pop-in point.

The stress a1c at pop-in and initial half crack length ac are used to calculate theGjc or Kic values. Even though the stable plane strain crack growth is subsequentlyarrested by the ability of the material to develop a shear fracture, the Kk value maybe computed from the mixture of the plane strain and plane stress (mixed mode)type fracture thaL develops as the crack extends to some critical length ac.

To summarize the above discussion, we find that the crack tip stress field is thedriving force for fracture, and the magnitude of this stress field is proportional tothe stress intensity factor K. K being a function of crack size, the applied stress andstructural geometry, it can be computed using structural stress analysis methods.The resistance to fracture is a material property defined as the fracture toughnessKc. Fracture occurs when K = Kc. Kc is also a ianction of the extent of plastic strain

Failure/

0

-J

.rPop-in"

Figure 17. Typical load-crack dis-

Oeflection (in.) placement curve.

=* I " L I 20

K I

I__ Figure 18. Transition -,f ',--#ure8, Thickmnes toughness.

15

". " '; . - . '.- -. -' -- ." . . " - - -.' . ' . - - . " .. - ." . ( ' , h% %

at the crack tip. If the plastic zone is small compared to the specimen dimensions,K, approaches a minimum constant value defined as the plane strain fracturetoughness, Kic. The relation between K, and KI, is shown in Figure 18 (Tobler andMcHenry 1983).

Fracture toughness test methodA standard test method, ASTM E399 (ASTM 1981), is used to measure plane

strain fracture toughness, Kic. The test is applicable to fatigue-cracked specimenshaving a thickness of 0.063 in. or greater in tension or a three-point bending load.Load vs displacement across the notch at the specimen edge is recorded graphi-cally. The load corresponding to a 2% increment of crack extension is established.The Kic value is calculated from this load by equations discussed earlier. The va-

a-: lidity of the determination of the KIc value depends upon the establishment of a"sharp crack" condition at the tip of the fatigue crack.

Fracture toughness tests for polymers and compositesIn the past, attempts have been made to characterize tlf : failure of nonmetals like

polymers, ceramics, rocks and composites using established fracture mechanicsparameters. However, to date no standard fracture toughness test procedure forthese materials has been established.

The complexity of fracture toughness tests of polymer materials arises primar-ily because of the large plastic zone at the crack tip and strain rate sensitivitytiviarshaii and Williams 1973). Most work in this area pertains to room tempera-

_ ture conditions, but low temperature studies on thermosetting epoxies (Kneifel 1979)and structural thermoplastics (polycarbonate) are available (Kneifel 1979, Martinand Garberich 1976, Tobler and McHenry 1983). Kk values for plastics are usuallylower than 10 MPa-m 1/2 at low temperatures.

Composites generally fail because of the development of flaws during fabrica-tion that become enlarged in service. At a microscopic scale this behavior is simi-lar to that of metal because metal crystals are known to be anisotropic, and randomaggregates of crystals are obviously heterogeneous. Thus, in a fracture mechanicsapproach, viewing the composites as statistically homogeneous leads to relationsbetween variables that are useful in engineering description and prediction of be-havior. Using this fracture mechanics concept, Corten (1972) has developed amathematical model for two-material, bonded composite fracture analysis. Al-though the concept offers a mathematical tool for the future development of compos-ite systems, the complexity of the problem has so far prevented this approach from

* becoming popular.

Influence of cyclic load on fracture toughnessWe have discussed that the basic material property essential to all fracture me-

chanics consideration is the material's fracture toughness, KI , at the temperature-. of interest. The Kk value gives an estimate of the critical defect size necessary for0 fracture with a single application of load. But in real life, few structures are built to

withstand only a single application of load. Consequently, the useful life of astructure depends upon the rate at which an existing defect will grow to the criticalflaw size as a result of fatigue (and also environmental) crack growth. Therefore,an additional material property basic to fracture mechanics consideration is therate at which a crack will grow under the loading conditions of interest.

16

0

N...%V~~~~~~~~~F A4,d - .. .. V*. ...

Clark and Wessel (1970), using their experimental results, have shown that therate o' -rack growth (da/dN) is exponentially related to the crack tip stress inten-sity per cycle (K), as given by

da - CO AK ndN

where a = crack lengthN = number of loading cycles

n,C0 = empirical constants determined from the test dataK = stress intensity factor per cycle.

Because the crack growth is influenced by cyclic loading, the recommended proce-dure for fracture toughness measurement includes consideration of fatigue crack-ing at nominal stress less than 25% of yield stress (Rolfe and Novak 1970).

Strain rate effect on fractureOne of the simplest ways to idealize viscoelastic material behavior under uni-

axial loading at an imposed strain rate dcidt is to express this behavior in the fol-lowing differential equation:

* 1 dcr a _ deE dt+ - (Maxwell solid) (13)o", E T dt

N, where E = Young's moduluscr = stresst = time

F_ = strainT, = steady-state tensile viscosity.

-. ' For finite increments of strain, stress and time, Ae, Aa and At, eq 13 can berewritten as

E - + A-- (14)

Atl 1E At

Now if time of loading At << 71/E, the term Aa/(rl/E) can be neglected, giving

A0 = E -AE (15)-. At At

which is a pure elastic loading. In this case brittle failure would result at the tip of acrack when the local stress exceeds the ideal tensile strength. Alternatively, if rI/E

* is much smaller than At, AalAt can be neglected, giving the equation of Newtonianviscous flow:

Ac E ". (16)71 /1E At

17

%.1%

T. . , L S 7- T

This equation shows that for small strain rates, there is always enough time for theelastic strains to relax-that is, the viscous deformation will alter the shape andorientation of the cracks, making brittle fracture less possible. Also, it follows thatbrittle fracture in glassy material (viscoelastic) is governed by the time constant ofloading.

%' Many investigators have examined the effects of strain rate and temperature onmaterial behavior (Duffy 1979, Klepaczko 1979, Nicholas 1981). For materials with

N strong temperature and strain rate dependence, the fracture toughness usually de-creases with decreasing temperature and increasing rate of loading. Fracturetoughness should therefore be evaluated in the region where it may show its mini-mal value and the conditions that reveal the transition from high to low values offracture toughness should be investigated.

In order to develop a suitable testing technique for high-strain-rate loading,Knott (1973) and Klepaczko (1979) have defined the loading rate parameter K1 toexpress how fast the crack tip region is loaded:

k',I= Kltc (17)

where KI, is the crack tip stress intensity factor in mode I (plane strain fracturetoughness) and t, is the time interval from the start of loading to the point when sta-

0 ble crack propagation starts.The spectrum of rate of loading as defined by Klepaczko falls into the following

categories:

I Quasi-static loading 1 MPa -5' 1 5 3 MPa 1'ss

(with a closed-loop

static dynamic testingmachine)

4 1 * -1II Instrumented hammer 10 MPa-m- s 5 k K!5 10 MPa #i-s

(for example, instrumentedCharpy hammer)

"- 6 -1 9 -1

III Stress wave loading 10 MPa f-m s < k, "0 MPa fms

Obviously, the KI spectrum covers about 10 orders of magnitude, and requires anumber of experimental techniques to be covered. For a K63 Al-Zn-Mg-Cu alloy,Klepaczko (1979) demonstrated that values of Kit measured at the stress wave load-

". ing are 20 to 36% lower than those for static loading, and for another aluminum al-Sloy (PA6) this value is almost half of that for static loading. However, Klepaczko's

[ •tests were all done at room temperature. Further reduction in fracture toughness isexpected at lower temperature.

.. 18

%-% %0%

'' " • =. "w " ,.% ".",% % % %' . o % % " -. .. ". . -". . . . . . . . . ." " " '" *"" * " -

TES'I PLAN

RationaleIn the past decade, fracture mechanics principles have been used to quantify

safety factors in structural design, taking into account crack propagation and/orbrittle fracture. Most structural members, components, vessels, and piping are de-signed according to analysis criteria that guard against failure, but variability inmanufacturing processes of materials may introduce defects or flaws. Despitenondestructive testing one must realize that flaws do escape detection on occasion

% and may grow to critical sizes under cyclic loading and stress corrosion cracking.It becomes important, then, to determine the associated minimum critical stressintensity factors, KI, that will lead to unstable growth in these cracks and also toknow the minimum stress intensity factor, KIa , at which these cracks will be ar-rested in a particular material.

Klepaczko (1979) has shown a significant difference between the static and dy-namic determination of Ki, at room temperature; at lower temperatures this differ-ence will possibly be higher, with Kic showing lower values. In cold regions num-erous field conditions may be cited in which dynamic loading on structures as-sumes critical importance. Wave action, wind action, ice sheet movement andimpact, drilling vibration on drill rigs, vehicular motion, blasting and earth-quakes are a few examples. The hulls of ice-breaking vessels are subjected to re-

* peated impact loading by ice, resulting in deterioration and the need for frequentrepair. Therefore, sound engineering design for cold regions should depend onmaterial property data determined at low temperature and at high strain rate load-

ing. The CRREL test program aims to generate these data. The next section dis-cusses this test plan.

Experimental programThe proposed testing system is summarized in Figure 19. For each material se-

lected to be tested, a number of test specimens will be prepared. The number willdepend on the expected variation in test results. Enough specimens will be made sothat statistically meaningful representative data can be obtained from each type oftest.

It is proposed that static, quasi-static and high-strain-rate tests be conducted toevaluate the fracture toughness property of each material. During its service life atlow temperature a component material is subjected to repeated low-stress-levelloading and occasional high-strain-rate loading. The testing should reflect asimilar loading pattern. Therefore, some specimens will be subjected to initial fa-

.S tigue loading by (tension-tension) load cycling. A comparison of fracture tough-ness data between the load-cycled and non-load-cycled specimens will reflect theeffect of servi:, life on degradation of material property.

r ,Quasi-static fracture toughness test* The quasi-static fracture toughness test (1 MPa 4i s-1 to 103 MPa 1m s-1) is pro-

posed to be carried out using modified ASTM compact tension (CT) specimens asdescribed by Klepaczko (1979). In this method, shown schematically in Figure 20, acompressive force is applied to a wedge-loaded specimen. The vertical displace-ment of the wedge is measured. Also, the transverse displacement of the specimenopening can be measured by a clip gauge. The measurements can be recorded by

19

-V

Ma-terial%saml

Temperature soaking230 to 270C0 to - 5C

- 200 to -250 C- 350 to -40 0C

Figure 1.test prora tceat

impt s - ~c~dr

Figurne 20.sur Deiefoausisaicsoaopesin fatlaigmhefrePatnnthed

wedgesis, applieb a tesetin mahie

Pemnn6aaBs

Mouu

Stent

I

two x-y recorders. in addition, an oscilloscope can record the force-time curvedriving fast crack propagation.

In the configuration of a wedge penetrating the specimen, the tensile force, Fc ,acting on the crack is given by

Fc (18)2 tan( .+

where a = half the wedge angle (degrees)= angle of wedge/specimen friction (degrees)

Pc = compressive force acting on the wedge (lbf).

And according to Klepaczko (1979), the fracture toughness KIc can then be denotedby

=j Pcfa/W) (psi G 7 (19)2B W-tan -+ 0

* 2)

where a = crack length (in.)W = width of the specimen (in.)B = thickness of the specimen (in.),

.(aIW) is given by

•1/2_ a/2 . 7/29/

f(a/w) = 29.6 a 185.5 + 55.7 - 1017 + 638.9 a

where B > (Kic/Gy)2 and ay denotes the yield limit of the material.

High-strain-rate testsThe high-strain-rate test is proposed to be carried out by using the split Hopkin-

son bar apparatus (Kolsky 1949). In this technique, a small cylindrical specimen is4 sandwiched between two long bars. A compression pulse generated by impact from

a third bar propagates down one of the bars and through the specimen into the secondbar. The bars remain elastic, although the specimen is deformed into the inelasticregion because of the impedance mismatch. The equations of one-dimensionalelastic wave propagation in long rods along with the recorded strain gauge signalsfrom the two bars are used to determine the time history of both force and displace-ment at the ends of the rods contacting the specimens.

In Klepaczko's (1979) stress wave loading technique, the geometry of the testspecimen is similar to that of an ASTM 399 fracture toughness compact tension(CT) test specimen with the pinholes cut off and an angular incision milled to ac-commodate a wedge for loading the specimen (Fig. 21). The wedge allows for

4 application of compressive force, whereas the crack tip remains loaded in tension.The displacement of the wedge, as a rigid body, denoted by UA is given by

21

r .,., . , ,,> = . € ;._ -- - :, 5 : ",' [ w . :, , . .. , .. ,.; .: ( ,o: . , ..... ..... ,._. ,. ,,,, .,_ .

1~

3fl . iu Il W V'- w -vv-PUNSs -~ vur -UWn -_V J. -- i-W-u-- ;.pY r~rw 'Kr 7' w i' -rWr W pww'r r w

Incident Bar I Tronsmit Bar 2

StrikerShockStritker Absorber

SolenoidTValve

"% X/At

Ccmpressor N2

"5 Pneumatic Hammer System Nitrogen Cooling System

',# Test Specimen

Figure 21. Split Hopkinson pressure bar applied to fracture dynamics.

UAt) =Co t It) - ER dt (20)

*and the displacement of the transmitter bar face, which is backing the specimen(denoted by u]), is

U Bt =CO f T(t) dt (21)

where EI , ER and ET are the elastic strains of the incident, reflected and transmittedpulses, respectively, and CO (in./s) is the elastic strain wave velocity in the bargiven by

Co = (22)

where E denotes Young's modulus (psi), g is the acceleration due to gravity (in./s 2 )and p is the density of the bar material (lb/in.3 ).

The net displacement of the wedge as a function of time is given oy

U(t) =u (t) - UB (t). (23)

Considering eq 20, 21 and 23, the ret displacement 8(t) (in.) is

8(t) =C O J i(t)- ER(t) dt - ET (t)dt. (24)

For conditions of equilibrium

-I (+ ER(t) = ET(t)-

o2

. . . . .

"L e; " : - : . ' - .? --." .- ": : -: 'i:; : -: -"::::% : "'-i :i ''::-i . ?.4'--- %.: i.:i l i: ? ~ ~ .: -": i : ., .:y : :: ---

lim W W: W W _ _-M- 6 , SS Wb31.1.WL WL L -VW t._ w ' .... ~ ,p - .W r .r.n. rr.p.-

4i Rewriting,

S 1 W(t) - eT (t) =- R (t). (25)

Substituting the above result in eq 24 above

8(t) = -2C 'eR(t)dt. (26)

If we assume that the bar cross-sectional area is A (in.), PA(t) = the instantaneousforce (Ib) at section A, and PB (t) (Ib) = the instantaneous force at section B, theforces acting on the specimen are the following:

4, PA(t) =EA i(t) + CR(t)] (27)

and

.PB (t) =EA [T (t)]. (28)

The average force P(t) on the loading system is

P(t) =- 1.A + PB (0) (29)

From eq 27 and 28o,

P _t EA [it+ Rt+ Tt].(30)

Again for equilibrium conditions, eq 25 is satisfied, and using this equation in eq30 we have

SP (t)= EA T (t). (31)

4.'

It will be seen that eq 28 and eq 31 are the same; therefore

SP(t) = PB (t) (32)

Equation 32 shows that the average force is proportional to the transmitted force,which can now be measured to determine the KI, value. If the critical point (pop-in)on the ET(t) trace can be detected, the critical force Pc can be calculated:

23

4-,

!0

P (tJ =EAETC(t (33)

where t, is the time period after which the crack starts to propagate. Once P, is de-termined, Kic can be calculated from

EA TC(tcf( a/W) (34),S{Ki, = psi. (4

2B f tan[c2) + tan -1 1.1

where a = wedge angle (degrees)W = width of specimen (in.)a = half crack length (in.)p = coefficient of friction between wedge and material surfaceB = thickness of specimen (in.).

Thus, if the coefficient of friction is known, the fracture toughness KIc may becalculated from the record of P(t) or P(8) where 8 is the wedge displacement.

Test equipmentThe complete range of testing from quasi-static (10-3 strain/s) to high (103

* strain/s) strain rate will require loading equipment, specimen clamping devicesin a variety of configurations, transducers, sensors, suitable recording instru-ments and signal analyzing devices. Table 1 lists this equipment, instrumenta-tion and facilities.

Table 1. Mechanical equipment and instrumentation require-ments for low temperature testing of materials.

_'ype of equipment Description

Mechanical 1. Universal testing machine with necessaryaccessories, 200 kips (intermittent use-MTS testing machine available at CRRFLwill suffice)

2. Fracture toughness and fatigue testingmachine, 20 kips (continuous use), includesload frame, actuator, hydraulic powersupply, load cells, strain transducers,grips and fixtures, environmental chamberand control console

3. Charpy testing machine

Instrumentation 1. Four-channel, 500-kHz digital oscilloscopewith signal storing, analyzing, plottingand computer interface capability

2. High resolution camera3. High resolution microscope

Fabrication material 1. Testing frame angle irons and miscellaneousfasteners

2. Impactor/gas system3. Electronic parts, strain gage and accessories

4. Pressure bars

24

o-0

k.

A large number of tests to determine primary elastic material properties at roomand low temperature will be conducted using test specimens in the MTS universal

' testing machine currently available at CRREL.A study of the effect of quasi-static strain rate on fracture toughness will be con-

ducted, both with pre- and post-fatigue conditioning of the specimens. Fatigue-test-ing, tensile testing and fracture toughness testing will be carried out using the 20-kip machine.

Composites and plastics cannot be tested for brittle fracture/toughness using the20-kip testing machine. For these tests it is necessary to use a Charpy testing ma-chine that can be remotely operated in a low temperature environment chamber.

High strain rate studies will be conducted in the Hopkinson bar test setup in-stalled at CRREL. The specimen will be jacketed for low temperature testing. Thetest setup will be designed for use of two 8-ft-long pressure bars, one 1-ft-long impactbar and an impacting mechanism using a gas gun and associated remote controlvalving. The instrumentation consists of the following:

1. Impactor velocity: photo-diode-activated timer.2. Triggering circuit electronics: accelerometer/vibration pickup.3. Longitudinal, torsional and flexural strain wave pickup strain gauges.4. Reaction pad instrumentation: dynamic load cell.5. Digital oscilloscope to record incident, transmitted, and reflected stress wave

pulses, and to integrate.6. Associated electronic parts for connection and interfacing of various equip-

ment.

EXPECTED INFORMATION

Expected information from the tests on each material will include its behaviordata, within both elastic and inelastic limits. Within elastic limits, studies will bemade of the change of modulus or stiffness with decreasing temperature and in-creasing strain rate. Conventional uniaxial tension, compression and flexuretests and Charpy V-notch test results will establish the baseline material charac-teristics at decreasing temperatures. For engineering design input, data will begenerated by fracture mechanics tests. Results of these tests will provide a stress

I.-. concentration factor at the crack tips, KIc values, and the minimum crack size thatcan be tolerated at various temperatures and loading rates.

Data from fatigue tests to failure at various low temperatures will be generated tocharacterize the materials. In order to reflect the conditions of failure of materials

* in service, fracture mechanics tests will be performed on specimens after subject-ing them to a given amount of cyclic loading

The initial emphasis of the testing will be on polymeric composites; standard testmethods are not available for these materials. Although testing of steel and otheriso tropic materials has achieved high sophistication through fracture mechanicstests, fracture mechanics of composites are not well understood. Like the material

• itself the art of composite material testing is also in a development phase, and theproposed test program will contribute to this evolutionary process.

Low temperature property data will be of great value to composite structuredesigners. Besides basic stiffness and strength data, several types of ancillarydata may be generated di ring the testing process. For example, acoustic emissions

* from the sample may b- studied during tensile loading of composite beams and re-

25

%

lated to its behavior through the fracture process. With infrared imagery tech-niques crack nucleation spots may be identified and crack growth may be fol-lowed. X-ray study may show the filament dislocations developing during testing.Fractography and scanning electron micrography techniques may be employed todevelop understanding of the fracture processes of composites. Room temperaturedata generated by these test methods are available but low temperature data aresparse. The low temperature data will contribute to the basic understanding of lowtemperature behavior.

At later dates, special steels and other materials for cold application will be in-cluded in the test program. Expected results from the tests of structural metals, e.g.

6 % steel, will be based on the approach that the standard fracture toughness testing is acomplex task in its present form in terms of specimen preparation, instrumenta-tion and recording of the data. This complexity increases more when the tests are tobe done on material at temperatures below or near their nil ductility temperatures.As a result, Charpy-V-notch impact testing (CV) is still the most frequently usedtest method, despite large scatter in the data. We plan to evaluate simple-geometryspecimens in dynamic tests such as the Hopkinson pressure bar test to obtain roomand low temperature fracture toughness data as a conservative estimate of thisproperty. The Hopkinson pressure bar method is simple and less time-consuming,and through proper data processing equipment, results are obtained quickly. Spec-imens can be cooled to their nil ductility temperatures by a refrigerant inside a

* jacketing around the specimen, and the test is finished in a short time because ofimpact loading. It is expected that a large volume of low temperature data can be

" - generated at much less cost by this method than by standard fracture mechanicstests.

Material specimens for tests will be obtained by contacting the Army researchand materials development communities that have the real need for such materialsdata. For example, a need for low temperature composite behavior data has already

*. been identified by the U.S. Army Materials Technology Laboratory, Watertown,Massachusetts, and it has provided the first batch of composite materials for test-ing. Results of the study will be published as CRREL reports and will be madeavailable to primary users (AMC, TRADOC) and other users (TACOM, MER-ADCOM, ARDC) through OCE (DAEN-ZCM). The results will also have a high po-

a- tential for non-mission-related technology transfer to local, state, and otherfederal government agencies as well as to private industry.

SELECTED BIBLIOGRAPHY

Argon, A.S. (1974) Statistical aspects of fracture. In Composite Materials: Fractureand Fatigue (L.J. Broutman, Ed.). New York: Academic Press, vol.5.Ashby, M.F. (1972) A first report on deformation-mechanism maps. Acta Metall.,20: 887-897.ASTM (1966) ASTM Standard, Philadelphia: American Society for Testing andMaterials, Part 30.

,'N. ASTM (1966) Behavior of materials at cryogenic temperatures. Philadephia:American Society for Testing and Materials, STP 387.ASTM (1981) Annual Book of ASTM Standards. Philadephia: American Society

"" a for Testing and Materials, E399, part 10, pp. 588-618.

S26

:'0,C

I1IX,

Baldy, M.F. (1976) The development of large-diameter high strength line pipe forlow temperature service. Offshore Technology Conference, Paper no. OTC2667.Berg, K.Rl. (1983) Composites components--most likely to succeed in vehicles. InProceedings, Sixth Conference on Fibrous Composites in Structural Design,AMMRC, Watertown, Mass., pp. 67-70.Boyle, R.W., A.M. Sullivan and J.M. Krafft (1962) Welding Journal (N.Y),Research Supplement, 41:4285-4325.Clark, W.G. and E.T. Wessel (1970) Application of fracture mechanics technologyto medium-strength steel. In Review of Developments in Plane Strain FractureToughness Testing (W.F. Brown, Ed.). ASTM Special Publication 463, pp.160-190.Corten, H.T. (1972) Fracture mechanics of composites. In Fracture: An Advanced

%: Treatise, (H. Leibowitz, Ed.). New York: Academic Press, vol. VII.Couch, W.P., J.E. Gagorcik and T.F. White (1983) Structural evaluation of ad-vanced composite application to hydrofoil strut/foil system. In Proceedings, SixthConference on Fibrous Composites in Structural Design, AMMRC, Watertown, pp.25-65.Daziel, D. (1983) High technology test bed: Where we have ben, where we are go-ing. In Proceedings of the Sixth Conference on Fibrous Composites in StructuralDesign, AMMRC, Watertown, Mass., pp. 23-24.Duffy, J. (1979) Testing techniques and material behavior at high rates of strain.

* ., In Procee ings, Conference on Mechanical Properties at High Rates of Strain, (J.0 Harding, Ed.). The Institute of Physics (London).•. "-Favor, R.J., D.N. Cidieon, H.J. Grover, JE. Hayes and G.lVL McClure (1961) In-

vestigation of fatigue behavior of certain alloys in the temperature range room to-423'F. WADD Technical Report 61-132, Defense Technical Information Center,Alexandria, Virginia.Foye, R.L. (1983) Status of the NASA/DOD composite flight service evaluation pro-grams. In Proceedings, Sixth Conference on Fibrous Composites in Structural De-sign, AMMRC, Watertown, Mass.

'e Frank, W.F.X. (1979) Spectroscopic analysis of high polymers for use at lowtemperatures. In Nonmetallic Materials and Composites at Low temperature (A.F.Clark, R.P. Reed and G. Hartwig, Ed.). New York: Plenum Press.Griffith, A.A. (1920) The phenomena of rupture and flow in solids. PhilosophicalTransactions, Royal Soc. (London), Series A, 221: 163-198.Hall, W.J. (1969) Evaluation of fracture tests and specimen preparation. In Frac-ture (H. Liebowitz, Ed.). New York: Academic Press, IV:1-44.Hallet, B. (1983) The breakdown of rock due to freezing: A theoretical model. InPermafrost: Fourth International Conference, Proceedings, Washington, Na-tional Academy Press, pp. 433-438.Hansen, O.). (1960) Materials for low temperature use: Short course on process

'V industry corrosion. In Proceedings, National Association of Corrosion Engineers,Houston, pp. 273-286.Harsem, 0. and H. Wintermark (1969) An evaluation of the Charpy impact test. InProceedings ASTM Annual Meeting, Symposium on Impact Testing of Metals, At-lantic City, New Jersey, NTIS No. N71-11585.Hartwig, G. (1979) Mechanical and electrical low temperature properties of highpolymers. In Nonmetallic Materials and Composites at Low Temperature (A.F.Clark, R.P. Reed and G. Hartwig, Ed.). New York: Plenum Press.Hartwig, G. and D. Evans (1982) Fundamentals and Applications of Nonmetallic

• Materials at Low Temperatures. New York: Plenum Press.

Z7

.%

Irwin, Gl. (1958) Fracture. Handbuch der Physik VI (FlUgg Ed.). Springer, pp.551-590.Irwin, Gi. (1960) Fracture testing of high strength sheet materials under condi-tions appropriate for stress analysis. U.S. Naval Research Laboratory, Report 5486.Kasen, M.B. (1975) Properties of filamentary reinforced composites at cryogenictemperatures. In Composite Reliability, Philadephia: ASTM STP 580, AmericanSociety for Testing and Materials, pp. 586-611.Kasen, M.B. (1983) Composites. In Materials at Low Temperatures (R.P. Reed andA.F. Clark, Ed.). Metals Park, Ohio: American Society for Metals, pp. 413-463.Klepaczko, J. (1979) Application of the split Hopkinson pressure bar to fracture dy-

r.. ~namics. Inst. Phys. Conf., Ser. 47., pp. 201-214., Kneifel, B. (1979) Fracture properties of epoxy resins at low temperatures. In Non-

metallic Materials and Composites at Low Temperatures, New York: PlenumPress, pp. 123-129.Knott, J.F. (1973) Fundamentals of Fracture Mechanics. London: Butterworth.Kolsky, H. (1949) An investigation of the mechanical properties of materials atvery high rates of loading. In Proceedings, Physics Society (London), vol. B62, pp.676-700.Kreibich, U.T., F. Lohse and I. Schmid (1979) Polymers in low temperature tech-nology. In Nonmetallic Materials at Low Temperatures (A.F. Clark, R.P. Reedand G. Hartwig, Ed.). New York: Plenum Press, pp. 1-30.

Lange, E.A. and F.J. Loss (1969) Dynamic tear energy-a practical performancecriterion for fracture resistance. Naval Research Laboratory, Washington, D.C.,NRL Report 6975.Ledbetter, H.M. (1979) Dynamic elastic modulus and internal friction in G-10CRand G-11CR fiberglass-cloth-epoxy laminates. NBSIR 79-1609, Boulder, Colorado:National Bureau of Standards.

4Marshall, G.P. and J.B. Williams (1973) The correlation of fracture data forPMMA. Journal of Material Science, 8: 138-140.Martin, G.C. and W.W. Gerberich (1976) Temperature effects on fatigue crack

growth in polycarbonate. Journal of Material Science, 11: 231-238.Martin, H.L., P.C. Miller, A.G. Ingram and J.E. Campbell (1968) Effects of lowtemperatures on the mechanical properties of structural metals. NASA Report SP-5012(01), Washington, D.C.McClintock, F.A. and S.A. Ali (1966) Mechanical Behavior of Materials. Reading,Mass.: Addison-Wesley, pp. 546-561.McHenry, H.I. (1983) Structural alloys. In Materials at Low Temperatures (A.F.

O Clark, R.P. Reed and G. Hartwig, Ed.). New York: Plenum Press, pp. 371-411.Nadai, A., (1950) Theory of Flow and Fracture of Solids. New York: McGraw-Hill.Nicholas, T. (1981) Impact Dynamics (J.A. Zukas et al., Ed.). New York: Wiley, p.227.Nichols, R.W. and A. Cowan (1969) Selection of materials and other aspects of de-sign against brittle fracture in large steel structures. In Fracture: An Advanced

* .Treatise (H. Liebowitz, Ed.). New York: Academic Press, V: 233-284.Parker, E.R. (1957) Brittle Behavior of Engineering Structures. New York: Wiley,

"4. pp. 253-301.Pellini, W.S. and P.P. Puzak (1963) Report No. 5920, Naval Research Laboratory,Washington, D.C.

* Plumer, JRL, Nil Schott, J.& McElman and S-X. Driscoll (1983) Design andfabrication of FRP truck trailer side rack. In Proceedings of the Sixth Conference

28;4"

L'. +' -"./ J- "."-* -•" -" ..- 'j.- -' - . *,, . . . ,.', .'.'..* . *.- :".. . - .. - -e-Je .% , .. C,fr-. % .... '

on Fibrous Composites in Structural Design, AMMRC, Watertown, Mass., pp. 23-24.Read, D.T. (1983) Mechanical properties. In Materials at Low Temperature (Reedand Clark, Ed.). Metals Park, Ohio: American Society for Metals, pp. 237-267.Reed, R.P. and R.P. Mikesell (1967) Low temperature mechanical properties ofcopper and selected alloys. National Bureau of Standards, Monograph 101.Rinebolt, J.A. and W.J. Harris, Jr. (1951) Effect of alloying elements on notchtoughness of pearlistic steels. Transactions, ASM, 43: 1175-1214.Rittenhouse, Jl. and J.B. Singletary (1968) Space Materials Handbook. ThirdEdition, Lockheed Missiles and Space Co. Technical Report AFML-TR-68-205. Air

Force Systems Command, Wright-Patterson Air Force Base, Ohio.Rolfe, S.T. and S.R. Novak (1970) Slow bend KI, testing of medium strength hightoughness steel. In Review of Development in Plane Strain Fracture ToughnessTesting (W.F. Brown, Ed.). ASTM Special Publication 463, pp. 124-159.Sparks, L.L. (1978) Cryogenic foam insulations: Polyurethane and polystyrene. InNonmetallic Materials and Composites at Low Temperatures (A.F. Clark, R.P.Reed and G. Hartwig, Ed.). New York: Plenum Press, pp. 165-205.Strawley, J.E. (1969) Plane strain fracture toughness. In Fracture: An Advanced

N Treatise (H. Liebowitz, Ed.). New York: Academic Press, vol. IV.Strawley, J.E. and W.F. Brown (1965) Fracture toughness testing methods. InFracture Toughness Testing and Its Applications, American Society for Testing

and Materials, ASTM Special Technical Publication no. 381, pp. 133-198.Tetelman, A.S. and A.J. McEvily, Jr. (1967) Fracture of Structural Materials.New York: Wiley, p. 268.

pTitus, J.B. (1967) Effect of low temperature (0 to 65 0F) on the properties of plastics.Plastics Technical Evaluation Center, Picatinny Arsenal, Dover, New Jersey,Plastic Report 30.Tobler, R-L. and H.I. McHenry (1983) i'racture mechanics. In Materials at LowTemperatures (R.P. Reed and A.F. Clark, Ed.). Metals Park, Ohio: American So-ciety for Metals, pp. 269-293.Tomin, M.J. (1983) Fracture mechanics approach to minimize risk of failure insteels for arctic use. In Proceedings, Fracture Toughness Evaluation of Steels forArctic Marine Use (R. Thomson and C.S. Champion, Ed.). Physical Met. Lab.Canmet., pp. 12-1-12-15.Witzell, W.E. and NR. Adsit (1969) Temperature effects on fracture, In Fracture:An Advanced Treatise (H. Liebowitz, Ed.). New York: Academic Press, IV: 69-112.

* Zinkham, R.E. and JRF. Dedrick (1969) Fracture behavior of aluminum alloys. InFracture: An Advanced Traise (H. Liebowilz, Ed.). New York: Academic Press,IV: 299-370.

6

29

!N

* APPENDIX A:~ LOW TEMPERATURE BEHAVIOROF SOME STRUCTURAL METALS AND ALLOYS.

800 1 1

Yed600

6002014 T65lIA 4 4Cu 0 8Mq 0 8Sn)

0u""

a"O_ 2219 T8512 400

400

n

6061 T651

Cn 5083 H4321

annealed200200

50?p,0

1099 4110010 0

00020 30 0000020300100 200 300

TEMPERATURE, K TEMPERATURE. K

Figure Al. Yield and ultimate tensile Figure A2. Yield strength of six aluminum*strengths as a function of temperature for alloys at temperatures between 4 and 300 K

annealed, commercially pure, and solution- (Kaufman et al. 1968) (in McHenry 1983).hardened aluminum and a precipitation-hardened alloy. Open symbols are yieldstrengths; closed symbols are ultimate tensilestrengths (in Read 1983).

o 2219-T6x 2219-T6 0 2219-T87

o2014-T60 a 2119-T6M

00

50 - 0

0 -

20 20

K 10S %' K 10.

504

~20 If2

-~ 0

-500 -. 00 -S00 -200 -100 0 00o 400 S00 200 00 0 '00

Temperature. F Temperature. F

Figure A3. Tensile properties of 2000-series Figure A4. Effect of heat treatment on 2219aluminum alloys in the T6 condition (Martin alunatn alloy (Martin et al. 1968).et al. 1968).

31z,

0%I,-Z

- -~ ~ ~ 3 '3-~ '~ - ~Mx .- ,l.w V1' -. 7 KM L-11KW W:Kf W WY -'-V-"-'~.~- .- ~ --. .w-wv 'WV' -J.u

x A356-T6 ___E E 9A612-F

79Q- - T

2'

%Temperawre, F Temperatur. F

*Figure A5. Tensile properties of 7000- Figure A6. Tensile properties of alumi-series aluminum alloys in the T6 con- num casting alloys (Martin et al. 1.9t)8).dition (Martin et al. 1968).

- I VIXUR61 '94 10. DIAL *0 @OIMPTIN F. HOMSIE 575119- 400 PSI; a a -% 20 ~CFO daD m PLASTIC,

0 S LAS? C. STIRM M GAI.ULAnD USIN*I ~G~ &NoI MW 3 Off TRUE 111 TWE PLASTI 1111,111.(

I&A

Figure A7. Fatigue behavior of 1100 aluminum (Favor et al. 1961).r%%IGELIE cce

K3

- LIM6 V4 a U OW MO OOWT* ?4.TCWILE STI SUS P09 R* -. S Wil

4 - - - - - --

4.4

10- - - - - - -

(Favo et--al--------

4990 PS. - -1.90 M -- PATQ- L T T.116,1 1111 C OL'O SN

#N w 11 111 11 PATUMNS

6-

- I - - - - - - -O

*8 #~ WE lcs

Figur A9 Faiu falr of 202 almiu (Fvo et al91

I3

I V

FL9X I; '/A iL Ol 1o0 COISI4TION TO,1T1ENSILI %TIN 44,800 PSi; 41* * OW CON;OY c l.L S N O T E . C I

5C - i

SI--

PP.

.. o-

a'

00II 0 - - - on

, TEMPERATURE.,-

Figure A10. Fatigue strength of 6061 aluminum(Favor et al. 1961).

-" 0IOU 1 0 - 1 . 1W CV1 4w , O g uSL& I i1 - L - ASTa altasi W]i LCULAT9S

To M$ C/ I & ug gAs PLASIC

i

FATIGUE LIFE. Cyci

Figure All. Fatigue behavior of 6061 aluminum (Favor et al. 1961)

34

0

PLUAI;V4 04. 60.CONDOION TO. TENS"LSTN SknO 1 11 Rl 1. 1 000 41 1; 010L99

#40 RolloO. 3I

4'.o

It

04 --

-40 -4 -a -4;4 40~ 400"

TEMPERATURE, -F

Figure A12 Fatigue strength of 7075 aluminum(Favor et al. 1961).

Me 0691dS" w .t"co$wmFFi LD c

IKATG 10901 WR AWA9 SN

1010N A O W90T9PATCNW

I I-

I SS

F0u_ A3 aiu ehvo f77 lmnu Fvre l 91%4

so5

toM I

as1040.0

0AIU IFcce

FiueA3 aiu eairo 05auiu Fvre l 91

200 A0E N

STA;NtESS STEEL X A553dleI

2000 - ,Ye'd0

15 8Wd. 300 GRAODE_vp 304 SOLU TION-ANNEALE 0

- ~~~~cold rolled 50% 2 AAIN TE

- i~ 1001, Tpe 304 aneae X

1000Tvo 304N 012 w1 %Ni

0

0 50

0 M0 20 300 0 100 200 300TEMPERATURE, K TEMPERATURE. K

Figure A14. Yield and ultimate tensile Figure A15. Temperature dependence ofstrengths as a function of temperature fracture toughness for various nona us-for zone-purified iron, 9-nickel steel, tenitic steels, showing the beneficial ef-and carbon steel. Open symbols indicate fects of increased nickel content onyield strengths; closed symbols indicate transition temperature and subtransi-ultimate strengths (in Read 1983). tion range toughness (in Tobler and

McHenry 1983).

300 I

AISI 31025

m 200-~200 -

1r0

II

4 00 10 300 -

0100 200 300

6TEMPERATURE. K TEMPERATURE. K

Figure A 16. Temperature dependence of Figure A17. Fracture toughness offracture toughness for alloys, illustrating 3.5, 5, and 9% N steels at tempera-characteristic behavior for three different tures between 4 and 300 K (Toblercrystal structures (in Tobler and Mc- et al. 1976) (in McHenry 1983).

* Henry 1983).

63

...................... %.-.

I OTATING *8AU; S/UO. PL.ATE4 WOMMULIEESSC -1x~ TE SIL T4: iIS0O PS11 a -I. 1750 RPM;

ft - -Room TEMP. EC31

- -UNNOTCHEO

0)So - -

FATIGUE LIFE, Cycles

Figure A18. Fatigue behavior of 2800 steel (Favor et al. 1961).

M lop - exuag; &A OI WAs 41A~iNIA~ff I MR,. Ala COOL.*DoorP ton. All OOL. ~Fo 2 NAR COOL.

ITELNSIL91 11TO 1.00 PSI, X .1 100 CFO;to- PLASTIC; ... o ELASTIC

VESS E CALCULATED USING SoIlI AND ARE

IGO- NOT TRUE IN iNS PLASTIC RANGE1. [a

a so-

FATIGUE LIFE, cycles

Figure 19. Fatigue behavior of 2800 steel (Favor et al. 1961).

37

lee6 - - F L ; al IAI . GA woo" I60 Is6.

AM COOL, 146o 0 NIL A 00WL F

tml AIR Ci OOL.. I9SIS:II I7*300ISO .. I1. RON OFl GY6L9$ ISOT1L 413

-', * I ~ va ni I

410

-------

%---

'fie-

-." TEMPERATURE, -F

Figure A20. Fatigue strength of 2800 steel (Favoret al. 1961).

U,8

.5

-U

.'

-U-U' " ""d"-

"" " " "

'" ": " "

",-

'' " """" "

" """2 , -.. - ' ,

"' ' ,

V AI-I-WV -V wV rj

I2

h~6.

40-

00

-~FATIGUE LIFE, CYtlS,

Figure A21. Fatigue behavior of 4340 steel.

FL(I~t~ mamin. NUT

5*c - -- ____ ~ ir TRET!. T S.I I ll £1

a.,moC~

I -J

o Carpenter 20-Cb Ane1eoz2oo-0tp 0

i2 -'Age. -Nrenea

00--

50 -

2 30

Kt i0 Kt - 1

1PC

I -_ ___ ___ ___ ____ ___ ___ __ ___ ____ ___ ___ ___ ___

0 Ag6aree

U200

2020Kt 1

5. 0 Ag 0adee

0 20 20

350 -c -303 20 -00 0 '-00 -0 300 D

Te0praue 'F20 Tepeatre

Fiur A25.__ Tenil proprtie of HP 9--5 iue 2.Tesl poetisoA 5

stee (Mri 20 Kl 198.ste10rtnel.16)

z %

0' 0' z

2000 1

BAR

304 11600

S1200

2000

CL CARBON S1'EEL(Fe 0 8%C) 1

CD 1000400

Zone purdied IRON qun9 e NIKE EL(ePered

0 n00 100 200 300 102030

TEMPERATURE, K TEMPERATURE. K

Figure A27. Yield and ultimate tensile Figure A28. Tensile and yield strengths ofstrengths as a function of temperature for three austenitic strainless steels (AISI typesannealed, nitrogen-strengthened, and 304, 310 and 316) at temperatures between 4Kcold-drawn stainless steel 304. Open sym- and 300K (in McHenry, 1983).bols-yield atengths; closed symbols-ulti matetensile strengths (in Read 1983).

400 o Annealed%

I I I IG Age hardened, .00 -

Wee

0 8STAINLESS STER 200

300 X0 0

z 4KA4 ~ 300-

0oo6 2 ~00 -.

. .00

0 2834 0 KI 10<4004 -

-~0 0

0 0

002 0

0 C?

0 I I I 9100 t0' 102 1O3 10

4 10 106 101 C'50 .0 30 0 0 0

CYCLES TO FAILURE Temperature, 'F

Figure A429. Fatigue test results for stain- Figure A30. Tensile properties of l8Ni-8Co-less steels 304L and 304LNV at 300, 78 and 5Mo maraging steel (Martin et al 1968).4K (in Read 1983). 4

<K -sea I

-ar 10 - - -0 - 0 111 m - 40

ISS11 111-44 PON ft -N S

a Ir

fits

FATIGUE ,FK - -

FiueA2-aiu bhvo fAS 04sanessel.Fvre l 91

'a---- 2

w PII- -- x2

TEMPERATURE, -F TEMPERATURE. *F-400 -300 -200 -100 0 100 200 300 400 500 -460-400 -300 -200 -'00 0 100 200 300 400 50

*6o IaOs 160

140 140

i120

120 120

,Aged, A'tchd Ailed, CArp, V

cold 0Oa_ Z6. 100 - . -- -

Notched

* W~ -Cold Oco." 60%,~ 0 Cold 0,o 26% 1o

W hoooled, Nofthed,

60 6

4040 ------ _ _ _

* A,,oe/.d Cost, Chopy Keyhowe

z', loed 'Aged[

*0 100 200 300 400 So0 0 100 200 300 400 500TEMPERATURE.*K TEMPERATURE, K

Figure A33. Strength of copper (Reed and Figure A34. Impact energy of copper (ReedMikesell 1967). and Mikesell 1967).

TEMPERATURE, -F TEMPERATURE, -F-460-400 -300 -200 -100 0 100 200 300 400 500 -400 -300 -200 -100 0 100 200 300 400 500160 32in0-

cold Wo,'ked * 1201 2ko, AIOOLWS Or CLASnoryT

- -- - Arvmi or fl -t

Aged'. OPHfT ~'20 24 -_

- 100 -20 - - - - - - --

O A,,,,oled

60 12- - -

in 0 0- + C eld D ia 4 %

20

TEMPERATURE, *K TMEAUE*

Figure A35. Hard ness of copper (Reed and Figure A36. Modulus of copper (Reed andMike sell 1967). 43 Mikesell 1967).

% e

w w w w w w w w w ww..h,~~~-u.r.~ -. . ru s -r rr- r- ,- -- s -,rW.-

TEMPERATURE,*F TEMPERATURE, P* 40 30 20-100 0 00 200 300 400 500 -400 -300 -200 -100 0 100 200 300 400 500

* 140 1 '-- -40 -

,-~ojo,4 'ord90 t,705,4.O' A o~led, Chorpy V

12C 1 -20 ----- ~

100 -- - - - - - .--. 100 - -.. .

- odWorked.?5%-NdV)0 - - - - - --7-So

* A.000l0, Notched

* ~ c 7N.- 0- 50,Anele

900 N6-0 Col Wok01%

100 200 60 040, e9/e [ /0 A/o od., V? -otc 0

and ~ ~ ~ ~ ~ ~ ~ ~ -01 Mieel 97.(Re ndMksll16)TEMPEATURE F1do eMPERTURE FnWdl

40 k3a0

\ ' - 3 A n e a e d C l a- , V

200 - 20 -

90 /, A heted6 0-I0, Cnn old

0 100 200 00 4rnol 000 50 10 20 00 40 0

(60320

* I 639inAnEIA ir- - -6M-D514CSldFDrown

0 20 0 0 0 0 0 0 0 0 0

1EPE2TRE 2KTM4RTUE

FiueA91 adeso0u-0bas Re iueA0.Mdlso uZ basand ikeell 967. (Red nd Mkes~l 167)

OS-15 - a'le

APPENDIX B: LOW TEMPERATURE BEHAVIORe OF SOME SELECTED POLYAMRS

Tensile Stregt

300 40

300% Modulus

200 13001000 200

-50 -25 0 32 17 115 150 212

TESTWhG TEMPERATIME (*F)

Figure Bl. Effect of testing temperature on stress-strain propertiesfor natural rubber (Rittenhouse 1968).

4000

100 300% Modulus

-50 -25 0 -32 77 115 150 212TESTWIG TEMPEATURE ('F)

Figure B2. Effect of testing temperature on stress-strain proper-ties for butyl rubber (Rittenhouse and Singletary 1968).

45

00

6F -6 F 1-. 5 , C", N - 20 C 6.MI 1"/MIN ,5 -

S x T 3 3

":i I O 1 I I i i i ! 1 1

0 OO 200 300 0 O 20 300 0 00 200 300 0 l00 200 300% ELONGATION % ELONGATION % ELONGATION % ELONGATION

Zn6-+0Cr - .C 6-_20*C 6t-r, - 10o, 'MIN - I o/MIN 5-,O"/MIN,

4 ~ - 4 - 04 4

-. - '.I- -W

0' r orU' ' ' ' ' ' o - ' ' ' ' '2

0 100 200 3YjO 1 OO 200 300 0 '00 200 300 0 100 200 300I % ELONGATION % ELONGATION % ELONGATION % ELONGATIONV.i

6.- r T6 - s -c 6-s-200IN5 -20"/ MIN S .

.. . 0 . 2 04,I 0

'00 20C , ,2 / -3 02 - 0 IN2- 01cI

I I0" . O I I 0

.. 0 '0 00 300 0 0 200 300 0 100 200 300 0 100 200 0oo

% ELONGATION % ELONGATION % ELONGATION % ELONGATION,

OOP ---

Figure B3. Stress-strain curves of plasticized PVC resin (Titus 1967).

Lowering temperature and increasing rate of loading have the same effect. The curve at -5 0C and at aOle loading rate of2O in. /mi. is the same as that of-200 C at a loading rate of I in. / in. Therefore, a 20-

fold increase in loading rate is equivalent to a 15 'C reduction in temperature on the tensile properties* of this material

* Decreasing temperature and increasing the rate of loading changes the curves from soft to weak, tosoft and tough, to hard and tough, to hard and strong, to hard and brittle.

49'6

'I Io'r ne, . %.- . - - - - - 4'- 9 4 . 4. 9- " ,% C_ - -*-.4 4.- %' '

'.000,000

900.000

S700.000

- 600,000U

500,000

.000mooo.00 400,000

(n

-6.000300.000

YIEL0

200,000zD0

.000 JTIMATE IO(000

1 I I 0-40 0 40 80 '20 160 -40 0 40 so 120 160

TEMPERATURE !F b TEMPERATURE,.F

12,000

U1 0,000

4 LA a'. 00

rz .000

w2 22 ~ T 6 4000

,.,- 0 2,000

0 I 0-40 0 40 so '20 160 -40 0 40 so 120 460

C TEMP.RATURE.eF 0 TEMPERATURE, F

Figure B4. Tensile and flexural properties of vinyl chloride (Saran 909) (Titus 1967).

47

% %U." " : : . :-. % -

..

..p

% 67000t

TEE /

TOT

- -0 FEP . ,e v F000-4000 1 -TL

-69Fon 500004000 T3F ?5FS--

, ...000w* ,<m'+ -. , *

-J 73F..3000 -

* z 000 2?

2000 212r 0 1 o 3

1000 212 F $RI PRCN

0 I I I I I I 1V0 50 100 150 200 250 300 350 400 450 500 550 600

STRAIN IPER CENT)

Figure B5. Tensile stress vs strain (to fracture) of TEE and FEP (Titus 1967).

I.OW

KIAoo00 TRANSITION POINT

I 00,000

100

-300 -10 -400 -50 0 so P0 IGO 00 0 500

TIwIATUnI, e0

Figure B6. Tensile properties vs temperature of Teflon FEP film(Titus 1967).

48

% %

. . ..- . ., +. -.-. . + .,. -.+ -,+ - ,,, - . . ' . _. , + .. .

Table BL Effect of crystallinity on Teflon TFE fluorocarbon resins (Titus 1967).

Crystallinity, %Property -O - -

50 55 60 65 70 75 80

TensileYield Strength, Kpsi( 1 ) 212 1.20 1.14 1.08 1.04 0.98 0.92 0.83

73 1.90 1.80 1.60 1.40 1.30 1.20 1.00-40 3.80 3.20 2.80 2.40 2.30 2.10 1.60

Elongation, %(l) 212 - 44 39 33 25 - -72 58 50 40 32 25 14 6

-40 40 35 30 25 20 10 8

Flexural Modulus (2 ) 212 26 33 41 49 57 64 72.Elasticity, Kpei 73 45 67 WO 110 132 154 176

-88 210 240 270 310 340 370 400

(1) ASTM D638-52T modified for test specimens 3/16" x 1/16"x 7/8" for low-andhigh-temperature tests. Crosshead speeds of 0. 02 in./min. (Baldwin) for lowtemperatures and 0.05 in./min. (Intron) for high temperatures.

(2) ASTM D790

49

Table B2. Mechaical propmties of FE 24nL

Temperature (0OF)

0073 0 -20 -40

Tensile strengt1h (ps.) 8,320 14,820 16,570 17.990

'2-Yield St rength (ps.) 8.320 14,820 16.570 17,990

*Elongzati~.n ('r) 250 23 15 14

Flexu r I Modulus (p. s. i.) 66.500 336. 300 426. 00v 463.000

Moisture; 0. 19 -0. 23%Data obtained in accordance with ASTM D) 759.

60-50 240

22

0

so. ~ 0 5o0 .00 .50 0 so 100TEMPERATURE. *C TEMPERATURE. *C

*Figure B7. Maximum Figure B8. Variation of modulustensile strength of nylon of rigidity in torsion (Q) for two11 (Titus 1967). typical grades of nylon 11.

TEMPERATURE. *

-40 -20 0 20 40 60 so 10014000

12000 - TIEWPCIATU.E, 'C-000-40 -20 0 20 40 40 0 0

j 00000

- 0 0 . 4 0 0 0

C 2 4300000 C

400~ '0000

6002400 40001

-P0 0 *000IS 20 9 - 0 0 0 *0 20 20

500

000

- - - * 000 - - - --- -- . - - - - - . - .

*-200

100-0.&. A.I .. .A.J 1~ 1j r .t

Table B3. Tensile impact of Zytel nylon resins (Titus 1967).

ZYTEL

PropWtv 101O) W rI7 31(2) 37(2) 42(l)

T ens d e i m pl a ct . ft 1l ) s q i n .

Dry: at 73 F 82 74 79 103 155at 0° F 93 80 73 90 105a -40' F 85 88 68 94 107

Conditioned to 50'% RH at 730 F 156 1(4 121 135 300at 0°F 113 90 73 96 118at -40 0 F 79 71 86 88 110

(1) Nvon 66

(2) Nylon 610

24

4 -

20 -

x 1 )

012 '2

50

o -4

0 -- i-so -40 0 40 so 120 160TEMPERtATURtE. *F

Figure B11. Effect of temperature of nylon 6/6 (Zytel 101)on Izod impact strength at 0.3 percent moiSture content(Ttus 1967).

22.000

20.000

16,.000

Z16.000I

12,000

VI

10.000

.- 0 IN MIN.

8.000 -20 IN./ MUN

6.000

.02 IN MIN I/MdI" I I I I /

MilN'-150 -100 -50 0 so l00 150

TEMPERATURE (C)

Figure B12. Yield stress vs temperatures for polysulfoneat various loading rates (Titus 1967).

240

% ~~200 -1,0

160 a

.200% STRAIN.1.. ~RATE i MIN.').600

S200% STRAIN RATE

11 20 -(MIN -1I1

IL

.,..-160 -120 -80 -40 0 40 so r20 IGO -160 -120 -0 -40 0 40 W10,:0 100

"'-TEMPERATURE (*C ) TEMPtERATURE. (*C)

w ".

'."Figure B13. Percent elongation vs temper- Figure B14. Rupture energy vs temperature""ature for polysulfone (Titus 1967). The co- for polysulfone (Titus 1967).

! • efficient of thermal expansion is 3.1IxlO-,5 in. /in..:. -- F over the temperature range -22 ° *to +300°T.

!52

% % %.

F. - -- -- I

CL:0

-60 -40 -20 0 20 40 60 80 100 t20 140

TEMPERATURE. C

Figure B15. Modulus of rigidity for a typical generalpolypropylene ASTM D1043 (Titus 1967).

4

GEEA PURPOSEL PUPS

0 L .. V, I .-- II so 0 50 100 ISO 200 250 z -SO 0 so 100 ISO 200 2SO

Figure B16. Tensile yield strength Figure B17. Flexural modulus vsof polypropylenes (Titus 1967). temperature of polypropylenes

10 (Titus 1967).

VSKSTOS ILLEO

0

4GEE PURPOSE

-o 0 50 100 ISO 200 250

TEMPERATURE.,

Figure B18. Flexural modulus vstemperature of polypropylenes

(Titus 1967).53

-% % - -

600 0

S400 1 1 4 GENERAL "UPOSEGEAN I:JERNAL PURPOSE 060

_ 2 I I !V I dI, 4MPACT GENRAL PURPOSEEEAPRPS

so-IMPPCTCT

600z MAT240

03

-4O I-

-A_

2006~30 60a2 - 10

40 60 120160 -400 40 120-0 0 4 40 0 40STEMPERATURE (F) TEMPERATURE (F) TEMPERATURE (F)

10 20

4, 14

x 21 12

hi GENERAL PURPOSE _j

£IMPACT

4 ~ 4

2 2'

0 0"-40 0 40 so 120 lEO -40 0 40 60 120 60

TEMPERATURE (F) TEMPERATURE (F)

*% Figure B19. Temperature vs mechanical properties for typical general-purpose and impactgrades of compression-molded polystyrenes (Titus 1967).

54

.00VS 0 1

44

T YRI 767TYRIL ?67

040 0 40 eo 120 160 -40 0 40 s0 '20 IS0

TEMPERATiRE *F b TEMPERATURE .F

20.000MAXIMUM FIBER STRESS7'0'.000 IN FLEXURE .PSI

600.000 TYRIL 76? :00-XPE'S

50i0016.000 -ZERLON 150

ZE0O N1.01 .000

400.000 ZRO S .0

MODULUS OF ELASTICITY. PSI 12.000 - TYRIL 767300.000

-40 0 40 so '20 060 -40 0 40 11O 120 ISO

C E MPERATURE . F If TEMPERATURE. F

Figure B20. Mechanical properties of polystyrene copolymers as a function of tempera-

ture (Tius 1967). Note. Tyrit: ac,-ylonitrile--styrene copolymer, Zerlon: methyl methacrylate -8tyrene copolymer.

55

%% %

-- - - _. --.- I/.'. - w~1V1%'' I I- -1 V- V-I-

10000--

9000- _ _ -

0-

5000- --

400' -, ,5

3000-- -

iNN 2000-- - -L

100

0 100 200 30 [K1

'4-. Figure B21. Young's modulus of several epoxy-resins vs temperature (Hardwig 1979).

'.

DOES NOT RUPTUR I0 .,/PLASTICIZED VINYL 90 OUROMETER AT TEMPERATURE

[14 LU VE Igor

r'.- %

EE LOW LENIW DENSITY4' w POLYETHYLENE L o OWYTYL

EVA 1Io 'I00 POYEHYEN

~~~~0 10 -' EVA -0 0 0€'s ,o IS- 100 1 4 t 0 4 0 O% 144Mw DENSITY10o 4 POLYETHYLENE

TEMPERAATUNE ,OC TEMPEiRATURIEo OF

Figure B22. Stiffness modulus vs temperature of Figure B23. Effect of temperature on Izod im-polyuethylene, polyethylene copolymers and vi- pact strengths of (Marlex) polyethylene (Titusnyl (Titus 1967). 1967).

I

40

200-

I001

"o60

0.4

_1' Figure B24. Modulus of polyethylene as a functionI of emperture(Titus 1967).

o4

I0*

04 HG DICUP502

lik0.911dnsiy oytyee DUCUP-dicryl:

-IM .40 0 4o 0 IGO No0 ago 300TKUPIPIAT UiiR, *F

Figure B2. Modulus of polyertylene--unedctionslnlo menitu PEtus 1967). DN-co

1ik04 99dniypoytyee DI CUP--i y

peroxide; TAC-triallyl cyanurate.

57

- - ,- --- ,- S IN ..

101

Sz

,d09

V( z

01

0

00-10jO -5 0 10 A

TEPRTUE*

'Il

70000

60000

7000

0-

-0040 -0 0 2 0 0 6 0VTRi 0930ERTUE

A U0 -

23.C

100 f* 600

0 50 100 -50no 20 .0 -2 0 2 40 60 8 10STRAN, %TEMPRATUE'a

50000

00-DENSITY. a DENSIT ELONGATION RATE w

'oo 10%I 200 TY030 1000%/Mi3000 --0

40 -0 0 2 0 6 S 0-0 00 290 40 AC 60 0

TEMPERATURE. *C TEMPERATURE, OC

Figure B29. Yield strength and elongations (by Figure B30. Tensile properties (ASTM D-412) vsASTM D-412) on compression moldings of poly- temperature for polyethylene. (Titus 1967).ethylenes having densities of 0.915, 0.930 and0.960 (and melt indexes of 1.8, 2.1 and 0.5, re-spectively) at strain rates of 100%/mmn. and1000%/mmn. (Titus 1967).

U .'59

'I,"

00:-7.1.

-WI

. lhi

I-01' - 20*C

-p00

Figure B31. Stres-strain curves at various

0 10 20 temperatures of low pressure polyethyleneDEFORMATiCo % (Titus 1967).

"N

, E .o 10,000

600 TYPE~ 60001: .- o 9000- = U Gooo I-

TYPE T "4,,o/ h°°1,i TPSo6000 TYPE 6000

STMRTU FTYPELa TYPE I

*U 4000

"400 TYP 4I3 20 10-,

%/

2000 2 20

hiJ

[eL

-4 . 0 40 s0 120 60 -40 0 40 s0 120 160

TEMPERATURE,*F TEMPERATURE ,'F

~500 500,000

z 40->40,0

* 400- 3 00,000-u TYPE Xm

W, 200- - 300,000 TP

TYPE~ ELW,00T3E1TR.F 2

0 0

I.TEMPERATURE, OF TMEAUE

Figure B32. Strength of comprenion-molded polyethylene at various temperatures (Tius 1967).

ODI;N

106

2

a

-00 -40 0 40 so IO too 200TEMPERATUI, OC

Figure B33. Modulus vs temperature-unfilled cross-linked acrylate copolymer DPD-6169 (Titus 1967). DPD-ethylene ethyl acrylate copolymer; DI CUP-dicumylpe-ri de; TAC-triallyl cyanurate.

400

200

@40

10

- 4

0.1 - 1of4 .CUYA

-00 -g0 0 so IGO 10 WO. 20 200rIURtUm~uI, .P

Figure B34. Modulus of ethylene acrylate copoly-mer as a function of temperature (Titus 1967).

61

% % 5%J

%0

0 0 3 5p * off

4 4

4.9. 2 ~~~~LAMINATED PLEXIGLAS 0UI -02 .05 AIAE LXGATEMPEATU DEGEES ENTIGADE EMPEAT MR-DEG ED C EXIGA 0

II 0 ---- _ _ A

(a) (b).2.015

Z. =tI

-14

w 3

60 45'0 26-20 1"

00

00

*, 0

,0

400,000

300,000101

- 5 040010,00000 29

UJ

-J30,000

20,00 0

18,000 LUCITE 140 "1"LOY"3 LUCITE" 29 :

0 , 2

TEIMPRATUNE[ - (*F)

Figure B36. Modulus of elasticity of acrylic (Lucite) vs temperature(ASTM D) 638-52T) (Titus 1967).

210,000 .I

24,000 NO LOAD SPEED22,000I~j 0.05 INCH PER MINUTE:-

22,000

4 6,000

0

- 14,000

< 12,000

0

" g,OO0

0

-94 -56 -22 14 so as 122 ,56 194

TEMPERATURE INoF

Figure B37. Flexural strength variation with temperature of castacrylic sheet (Titus 1967).

63

.. . ..- -... . ......... .. ... ........ ., .. -

-~

40000

1 0,000 -4p

Z 000

-S6,000- -QO(

iz,4.000 - 0O°

' ?.000 I qrI,0 'I 1 I t

S -60 -20 0 20 60 100 140 10 220 2,0 -60 -20 0 20 60 100 140 I9O 220 260

TEMPERAtURE. *F TEMPERATURE, 'F

Figure B38. Tensile strength vs temperature of Figure B39. Compression strength vs tempera-polycarbonate molding material (Titus 1967). ture of polycarbonate (Merlon) (Titus 1967).

001

4

45 a

30 S25U -

20 -. 4

15, 4

"'"~ . 20 • -

0aN5

' 401

,':o-1 0 SO w 21 12 120 I19 I0 ,0 0 so 1 IS

- I

Flre QUIEPOCY.*F TIMIElATURI, ,ot

•Figure B40. Effect of temperature Figure B41. Torsional modulus of polycar-.. on modulus of elasticity of poly- bonate vs temprature (Titus 1967).L". carbonate resin (Tius 1967).

%m%

,%,3

0?0

1 4 z

12 - 500-

2 - 400

I ~ 3006 -

I " 200

1002

0 0 L n2=-100 0 00 200 300 75 60 40 20 0 -20 -40 -60 -00 - 100

TEMPERATURE. IF TEMP RATURE IF

Figure B42. Izod impact strength vs Figure B43. Urethane foam spring rate vs tern-temperature of (Merlon) polycarbon- perature (Titus 1967).ate (ASTM) D-256) (Titus 1967).

40

35-

30 -1

-401 F T*

20 -'

a6 -

0- + 1651F

4

5-2

0C . 4 L i I I 1J~ 00 000 000 00 10 20 30 40 50

TEMPERATURE IF STRAIN, PERCENT

Figure B44. Urethane foam yield point vs tern- Figure B45. Compressive stress-perature (Titus 1967). strain of low density density poly-

ethylene foam at various temper-atures (Titus 1967).

65

0 0

o

25

m 1

00

0

I '. I5

75 60 40 20 0 -20 -40 -60 -80 -100

TEMPERATuRE IF

Figure B46. Urethane foam absorbed energy vstemperature (Titus 1967).

0O5LG/CU FT 0 7 LU/Cu PT

20 20I

. STRAIN {INIIN) b ST RAIN (ININ}

40 60 4

IS LU/C FTC T 0

-@0O P

so, 30 /20" - 2o

10 20 10 15

i"

0% I

0 0

0 2 3 4 5a 0 1 2 3 4 5 6 7

lf CSTRAIN (IN/INbST AI OW N

V, Figure B47. Static stress vs strain curves for resilient expanded polystyrene foam (Resilo-Pak)

, at temperature extremes (0.5, 0. 75 and 1.3 lb Ifft3 densities) (Titus 1967).so

,eSF

50 -- 8 *

TEMPERATURE. K0 so 100 is@ 206 256 300

2511 2

225 PU -9 locoJ 60

206 (Lfernisaet 0i..197?1)40

01I)175 -*** 1200

1125(Caaphel at 0i.111551

PU(J.)P ~4Uthg"a125 (Cspe/*i. 52

S8O

250 Fiur B48.)" Sher trngh a afuc

__________________________foms it dnsiyprces,6ndon

pu(.L) A -41978).-

25 200 Figure B48. Young'strmodthus a n-

____F __)_ __-_tion___of temperature for PU and PSm

0i. L* (Carr 1 R 18 1 * o with density, proces, and oi-TUUIRTUI. -N -0 21 10 0 soetation as parameters (Sparks,7)

TEMPERATURE. K

TEMPERATURE. K0 so 106 IS6 206 250 300

260 1 1600

240 ]PS~i)p.5.5rg/,Jextruded -1100

2206

136 PUti/.IpA54.5k/.J2. (A'vidses at ci. 1973) 10

160- /PUt//h. A-54.59/rjv mca (CAPwphali at ai., 1552) 1

.140-600

120 36

tSoo

I PS.-' '51.3kg/rnomedad/Arvidaaa at a.1..173) f66 03

320

(Stawart at 81_ 1574) PIJJ4,O 3Ukg/iu(Car,. 19go)

-41 -406 -360 -261 -106 6 36

TEMPENATUHE.*f

Figure B50. Compressive strength as a function Of temper-A ature for PU and PS foams with density, process and orien-

tation as parameters (Sparks 1978).

66

U.S GVRNET RNTN FFC: 98-50-5-%22