doi: 10.1595/147106706x154198 high-temperature mechanical ... · fig. 1 schematic diagram of...

158

Due to their excellent chemical stability, oxida-tion resistance, and resistance to the action ofmany molten oxides, the platinum group metals(pgms): iridium, platinum and rhodium are widelyused for high-temperature applications involvingsimultaneous chemical attack and mechanical load-ing (1). Although iridium is more sensitive tooxidation than platinum or rhodium, it is the mostchemically resistant of all metals. Its resistance toattack by stable oxide melts is maintained up totemperatures above 2000ºC.

The melting point of iridium (2454ºC) (2) andits high strength even at temperatures above2000ºC make it a particularly suitable material forapplications under extreme thermal and mechani-cal conditions which preclude the use of platinumalloys or rhodium. Important applications of iridi-um and iridium alloys are as crucibles for pullingsingle crystals (e.g. yttrium-aluminium garnet(YAG)) and components for manufacturing andprocessing high-melting special glasses.

A knowledge of the high-temperature proper-ties of a material, for instance stress-rupturestrength and creep behaviour, is crucially impor-tant for the design of components used at hightemperatures. The current investigation is part ofan extensive test programme focused on the deter-mination of the high-temperature mechanicalproperties of the pgms, such as the stress-rupturestrength, creep behaviour (3) and elastic properties(4). In this work new investigations into the high-temperature properties of iridium are presented forthe temperature range between 1650ºC and2300ºC. The results are discussed in conjunctionwith data determined from earlier studies (3).

Methodology for Stress-Ruptureand High-Temperature Tensile Tests

The stress-rupture strength and the creepbehaviour of pure iridium and iridium alloys weredetermined with a testing facility developed at theUniversity of Applied Sciences Jena. The testing

Platinum Metals Rev., 2006, 50, (4), 158–170

DOI: 10.1595/147106706X154198

High-Temperature MechanicalProperties of the Platinum Group MetalsPROPERTIES OF PURE IRIDIUM AT HIGH TEMPERATURE

By R. Weiland and D. F. Lupton*Engineered Materials Division, W. C. Heraeus GmbH, Hanau, Germany; *E-mail: [email protected]

B. Fischer, J. Merker and C. ScheckenbachDepartment SciTec, Precision-Optics-Materials-Environment, University of Applied Sciences Jena, Germany

and J. WitteMelting Technology, SCHOTT Glas, Mainz, Germany

In order to provide reliable data on the high-temperature deformation behaviour of iridium,the high-temperature material properties such as stress-rupture strength, high-temperaturetensile strength and creep behaviour are determined for pure iridium in the temperature range1650–2300ºC. Analyses of the stress-rupture curves and the creep behaviour of pure iridiumsamples at 1650ºC, 1800ºC and 2000ºC imply that the fracture behaviour is controlled by twodifferent fracture mechanisms depending on test conditions, in particular applied load andtest temperature. The existence of the different fracture modes is confirmed by SEM examinationof the fracture surface of samples ruptured at high temperatures. Anomalies in the creep curvesand the results of high-temperature tensile tests indicate that dynamic recrystallisationplays an important role in the high-temperature deformation behaviour of pure iridium.

device, for the measurement of high-temperaturematerial properties up to 3000ºC, is shownschematically in Figure 1. It consists of a gas-tightspecimen chamber which permits investigationseither in air or under a protective gas atmosphere.In the case of iridium and iridium alloys, a gas mix-ture of argon with 5 vol.% H2 was used to protectthe material from oxidation and thus avoid areduction in cross-section of the sample due toevaporation of volatile oxides (5, 6).

The load can be applied in two different ways.For the constant-load stress-rupture experimentsthe load is applied via a steel pull-rod by means ofcalibrated weights. For the high-temperature ten-sile tests the specimen chamber is mounted in acommercial servomotor-driven test machine andthe steel pull-rod is connected to the load cell atthe crosshead of the test machine. This allows acontrolled variation of the applied load. Non-stan-dard specimens (with typical dimensions of 120mm × 4 mm × 1 mm) were used for all measure-ments. The samples were laser cut from hot rolledsheet material. The sample orientation was chosenparallel to the rolling direction.

Direct electrical heating achieves high heatingand cooling rates for the samples. The ohmic heat-ing method allows easy access to the sample, andgenerally straightforward operation.

The temperature is measured by a non-contact-

ing technique using a digital pyrometer(INFRATHERM IS10). The infrared pyrometerhas a small measurement spot (approximately 0.5mm in diameter). Due to the ohmic heating thehighest temperatures are found in the central partof the sample. This region is therefore scannedcontinuously by the pyrometer via a tilting mirror.By storing the maximum value of emitted radia-tion, the maximum temperature at the surface ofthe sample may be determined. This value is usedto adjust the heating current via a thyristor regula-tor connected to the primary winding of a 100kVA transformer. The sample, short-circuitedacross the secondary winding of the transformer, isheated by alternating current at 50 Hz. Over a zone30 mm in length around the centre of the samplethe temperature usually does not vary by morethan ± 5ºC. Once “necking” occurs in the sample,the temperature outside the necking regiondecreases, whereas the temperature within thenecking region remains constant at the intendedvalue. The design of the equipment thus guaran-tees uniformity of temperature throughout theduration of the test, despite the sample deformation.

The strain is measured with a non-contactingvideo extensometer consisting of a 17 mm chargecoupled device (CCD) camera with 1280 × 1024pixel resolution. A special arrangement of telecen-tric lenses allows only near-parallel rays to pass the

Platinum Metals Rev., 2006, 50, (4) 159

Fig. 1 Schematicdiagram of equip-ment forhigh-temperaturecreep measurements


aperture, thus minimising perspective distortionscaused by variations in the distance to the object.Both the CCD camera and frame grabber are con-trolled by “SuperCreep” software, developed atthe University of Applied Sciences Jena, whichuses digital image analysis.

As mentioned above, ohmic heating causes thehighest temperature to be limited to the centralpart of the sample, to which creep deformation isnormally also limited. Strain at this part of thesample is determined by “SuperCreep” from con-tinuous measurements of the distance betweentwo markers. Suitable markers for high-tempera-ture tests on sheet materials are made by lasermachining samples of the material with four smallshoulders (Figure 2). The distance between thetwo corresponding markers on the same side ofthe sample is 10 mm. Since the part of the samplebetween the markers experiences a uniform tem-perature, the exactness of strain measurements canbe guaranteed, without their being influenced bythe temperature gradients near the ends of thesample.

A detailed description of the testing facility andthe algorithm for strain measurement is given in(7) and (8).

Material PreparationThe iridium raw material was melted inductive-

ly at 2550ºC in air in a zirconia crucible. Afterelemental analysis the ingot was forged at temper-atures between 1400ºC and 1600ºC. The forged

material was then hot rolled at moderate tempera-ture to 1 mm thick sheet. The iridium sheet wasfinally subjected to a special annealing procedureso as to recrystallise the deformed material withoutsignificant grain growth.

Scanning Secondary Ion MassSpectrometry (Scanning SIMS)

The microanalytical investigations were per-formed with a Cameca IMS 4f-E6 scanningsecondary ion mass spectrometer. Secondary ionmass spectrometry (SIMS) allows the detection ofvery small amounts of impurity elements in thematrix. Since both the species of detectable sec-ondary ions and their detection limits differ asbetween the positive and negative secondary ionspectra, different primary ions were chosen for theexcitation of the secondary ions. Oxygen primaryions were used for the investigation of the positivesecondary ion spectrum emitted by the iridiumsamples. The emission of the negative spectrumwas induced by caesium primary ions. It couldthus be ensured that all possible impurity elementscontained in the iridium samples were detected.

Metallographically prepared samples were usedfor the scanning SIMS investigations. So as to beable to investigate impurity levels both inside thegrains and at the grain boundaries, areas of thesamples containing grain boundaries were chosen.

It should be mentioned that the intensity of theemitted secondary ion spectrum is dependent onthe crystallographic orientation of the grains.Thus, if several grains with different orientationsare contained in the area under investigation, thiswill be indicated by differences in the brightnessattributable to the respective grains due to differ-ences in ionic emissivity. Thus grain boundariesmay be identified in the secondary ion spectrum,even if they do not contain significant amounts ofimpurities.

Stress-Rupture Strength ResultsThe stress-rupture strength of pure iridium was

determined in the temperature range1650–2300ºC. The results of these investigationsare summarised in Figure 3. The present experi-ments showed an excellent degree of

Fig. 2 Image of a creep sample with markers for thevideo extensometer (7)


reproducibility; the results are in good agreementwith those of corresponding measurements forshorter testing times reported earlier (3). In con-trast to the conclusion in (3), additional data onstress-rupture strength obtained recently, togetherwith a detailed analysis of the rupture behaviour(see below under ‘Fracture Behaviour Results’), ledto the conclusion that the stress-rupture data canbest be approximated by two intersecting lines.The discontinuity in the slope of the stress-rupturecurves correlates very well with a change in thefracture behaviour of the samples examined.Under high loads pure iridium shows a ductilefracture mode, whereas under low loads and longtimes to rupture, iridium tends towards brittleintercrystalline fracture. In fact, the discontinuitydoes not usually occur as sharply as indicated inFigure 3.

Samples taken from near the discontinuityoften show mixed fracture modes, partly intercrys-talline and partly transcrystalline. Since insufficientstress-rupture data in the range of the transition areavailable, the stress-rupture curves for the temper-atures between 1650ºC and 2000ºC areapproximated for the sake of simplicity in the man-ner shown in Figure 3. In particular, thestress-rupture curve at 2000ºC shows a very pro-nounced change in the slope at loads between 4and 5 MPa. This leads to strongly reduced stress-

rupture strength values at testing times longer than300 h. It is not yet clear whether this steep decreasein slope can be attributed solely to the change infracture mechanism, or whether the effect ofweakening of the grain boundary coherence isenhanced by very small amounts of impuritiesaccumulating at the grain boundaries after longtesting times at high temperatures. As reportedbelow, secondary ion mass spectrometric investi-gations showed that the impurity content in theiridium samples examined is very low.Nevertheless, it cannot be excluded that even verysmall amounts of impurity elements accumulatingat the grain boundaries can have a detrimentaleffect on grain boundary coherence, thus leadingto significantly reduced times to rupture.

Measurements at 1650ºC and 1800ºC have beenperformed up to approximately 1000 h duration.Extrapolations to testing times longer than 10,000h may not be considered meaningful. Data on thestress-rupture behaviour of pure iridium at 2200ºCand 2300ºC are available up to times to rupture ofapproximately 500 h and 150 h, respectively, andnot as yet for longer times to rupture. Within theavailable range the stress rupture curves do notshow a visible discontinuity in slope. Because ofthe very distinct decrease in slope in the stress-rup-ture curve at 2000ºC under low loads, noextrapolations beyond the measured times to rup-

Fig. 3 Stress-rupturestrength ofpure iridiumin the temper-ature rangebetween1650ºC and2300ºC

ture have been performed for the data at 2200ºCand 2300ºC.

The interpolated and extrapolated data on thestress-rupture strength of pure iridium at differenttemperatures are given in Table I.

Creep Behaviour ResultsThe investigation of the high-temperature

deformation behaviour of iridium revealed that,depending on test temperature and load, its creepbehaviour can be described by two types of creepcurve which differ significantly in shape.

Particularly at the lowest test temperature of1650ºC, under moderate load, the creep behaviouris represented by typical creep curves such as thosein Figures 4(a) and 4(b). These figures representthe creep behaviour of pure iridium at 1650ºCunder a constant load of 13 MPa exhibiting thethree well known stages of creep – primary, sec-ondary (or steady-state), and tertiary – frequentlyreported in the literature.

Under higher load, and particularly at highertest temperatures, the creep curves of iridiumshow significant anomalies. In the range of steady-state creep the creep curve contains differentplateaus, as shown in Figure 5. These plateaux areseparated by an acceleration of the elongation.This acceleration of creep is clearly visible inFigures 5(b) and 5(d). The phenomenon may becaused by dynamic recrystallisation, as was report-ed in (9) and (10), whose authors obtained creepcurves of similar shape when investigating thecreep behaviour of lead and copper, respectively.It can be seen in Figures 5(c) and 5(d) that in somecases more than one discontinuity occurs in thesecondary creep stage. This indicates that dynamicrecrystallisation takes place successively severaltimes during the secondary stage of creep. This iscalled periodic or cyclic creep (11).

These accelerations of creep complicate thedetermination of a constant creep rate in the sec-ondary creep range. The creep rate has thereforebeen calculated as an average for a time rangefrom 10% to 90% of the period of measurement.Thus additional contributions to the average creeprate from the accelerated creep in the transientcreep stages are included in the values. This aver-


Table I

Stress-Rupture Strength of Pure Iridium at Various Temperatures

Time to Stress-rupture strength, MParupture, h

1650ºC 1800ºC 2000ºC 2200ºC 2300ºC

1 31.8 24.4 14.1 7.1 5.410 27.7 18.4 8.9 4.4 3.3

100 15.6 11.0 4.6 2.7 2.01000 8.8 7.0 1.5 – –

10,000 5.0 4.4 – – –

Fig. 4 Creep curves: (a) of pure iridium at 1650ºC undera constant load of 13 MPa in Ar/H2 atmosphere, and (b)corresponding creep rate as a function of time. The meancreep rate for the period 16.95–152.54 h is 1.3 × 10–7 s–1


age creep rate has been used instead of the mini-mum creep rate for the calculation of the Nortonplot (Figure 6). In the Norton plot, the averagecreep rate for pure iridium determined in this wayfor each test temperature is shown as a function ofthe initial applied stress on a double logarithmicscale.

For the calculation of approximate trend lines,

the stress dependence of the average creep rate hasbeen assumed to obey a power law (Equation (i)):

dε/dt = f (S, T) σ n (i)

where ε is the elongation and f (S, T) is a functionof the structure of the material and of the temper-ature. For the isothermal representation in theNorton diagram, f(S, T) has been assumed con-

Fig. 5 Elongation and creep rate of pure iridium as a function of time in Ar/H2 atmosphere at 1800ºC for loads of:(a) and (b) 13 MPa. The mean creep rate at a constant load of 13 MPa for the period 5.68–51.12 h is 1.0 × 10–6 s–1.(c) and (d) 9.5 MPa. The mean creep rate at a constant load of 9.5 MPa for the period 22.5–225 h is 2.6 × 10–7 s–1

Fig. 6 Averagestationary creeprate of pureiridium as afunction of theapplied initialstress for tem-peraturesbetween 1650ºCand 2300ºC(Norton plots)


stant. The effects of structural changes, forinstance, due to recrystallisation, have beenaddressed to some extent by the use of the averagecreep rate, determined as described above. TheNorton exponent, n, contains information aboutthe nature of the prevalent creep mechanism in thesample. For creep mechanisms that are based onlyon the diffusional transport of material due tovacancy gradients either inside the grains (12, 13)or along the grain boundaries, a nearly linear stressdependence of the creep rate will be obtained, andn will be close to or equal to unity. For creepprocesses that are determined mainly by diffusion-controlled dislocation climb, n falls between 3 and5 for many common pure metals. In rare cases, nvalues up to 11 have been found.

The distinction between the different creepmechanisms under low and high loads mentionedin the preceding section must also be taken intoaccount in the Norton plots. The data for each testtemperature between 1650ºC and 2000ºC aretherefore approximated by two intersectingstraight lines. As already explained for the stress-strain diagram, the information available for2200ºC and 2300ºC does not allow this distinctionto be made for these temperatures.

For test temperatures between 1650ºC and2000ºC, the values obtained for the Norton expo-nent under low loads n1 fall in the range between 4and 5.5, and are thus in good agreement with theabove-mentioned values of the Norton exponentfor diffusion-controlled dislocation climb in manyother pure metals. For conditions under whichiridium exhibited a ductile fracture mode, i.e.under high loads, the Norton exponent n2 took

considerably higher values (5.7 < n2 < 13.7), indi-cating that under these test conditionsdiffusion-controlled dislocation climb is not theprevalent mechanism of deformation. The valuesfor f(S,T), n1 and n2 determined by approximationof the experimental data in the Norton plot usingEquation (i) are listed in Table II.

The temperature dependence of the stationarycreep rate can be expressed by an Arrhenius term.Thus, Equation (i) can be rewritten in the form(Equation (ii)):

dε/dt = ασ n exp–(QC/RT) (ii)

where α is a factor dependent on structure, QC isthe activation energy for creep, R is the gas con-stant and T is the temperature in K.

The activation energy, QC, for the creep mech-anism can be obtained by plotting ln(dε/dt) versus1/RT. If QC is independent of temperature,ln(dε/dt) will show a linear dependence on 1/RT,with the slope of the straight line equal to the acti-vation energy QC. In the present investigations itwas not possible to determine QC in the waydescribed, as the creep tests were performed underconstant load, not under constant stress. In thiscase the true stress is a function of the elongation,thus invalidating this method for determining theactivation energy.

Fracture Behaviour ResultsThe iridium samples showed excellent ductility

at the temperatures investigated. Particularly athigh loads, rupture strain values up to 100% havebeen measured. At lower loads, the values for therupture strain proved to be considerably smaller.

Table II

Coefficients for the Approximation of the Quasi-Stationary Creep Rate as aFunction of Stress According to Equation (i)

Temperature, f(S,T) f1(S,T) f2(S,T) Norton exponentsºC n n1 n2

1650 – 1.3 × 10–13 4.5 × 10–25 – 5.35 13.68

1800 – 1.2 × 10–11 4.4 × 10–16 – 4.40 8.28

2000 – 5.1 × 10–10 3.6 × 10–11 – 4.11 5.70

2200 2.5 × 10–9 – – 5.38 – –

2300 2.1 × 10–8 – – 4.99 – –


This finding is in accordance with the observationthat iridium exhibits two different fracture modesin stress-rupture tests at high temperatures.According to the literature (9) this change in frac-ture mode should be accompanied by adiscontinuity in the slope of the log-log plot in thestress-rupture diagram (Figure 3).

Whereas intercrystalline fracture occurs withoutnecking at low loads (Figures 7(a) and 7(b)),together with the appearance of extensive inter-crystalline crack formation across large parts of thesample (Figures 7(c) and 7(d)), the fracture at highstresses occurs by the transcrystalline mode,accompanied by extensive necking as shown inFigures 8(a) to 8(c). This is in accordance with text-book reports (11) of creep crack behaviour at hightemperatures for many materials which are proneto brittle intercrystalline fracture under low loads,but tend to transcrystalline fracture under highloads. It should be mentioned that the numerouscracks and fissures occurring along the grainboundaries during the creep tests under low loads

make an additional contribution to the elongationof the sample. This apparent ductility does not,however, influence the intrinsically brittle mode offracture. Samples exposed to high stresses exhibit-ed only very few intercrystalline cracks. As aconsequence the high apparent ductility of thosesamples can be assumed to be identical to the trueductility of the material.

On the surface of the samples exhibiting ductilefracture behaviour, slip bands that have formed inthe necked region are clearly visible. These are verydistinct, and in some cases extend across grainboundaries, sometimes covering several grains.Overlapping slip bands in different directions wereobserved in some grains. This indicates that differ-ent slip systems have been activated during thecreep experiments. It is not yet clear whether theslip bands are formed only in the final stage ofcreep deformation, when the load-carrying cross-section of the sample is significantly reduced dueto progressive strain and the true stress and creeprate increase rapidly, or whether the slip bands are

Fig. 7 SEMimages of the sur-face area near thefracture for pureiridium afterstress-rupture testat 1800ºC underan initial load of6.7 MPa: (a) and(c) ×20; (b) ×100;(d) ×50


formed throughout the creep experiment.

High-Temperature Tensile TestsThese tests were carried out in order to exam-

ine the behaviour of pure iridium at hightemperatures under dynamic loading. Figure 9

summarises the results for the temperature rangebetween 1600ºC and 2300ºC. The shapes of thestress-strain curves, particularly those for temper-atures between 1800ºC and 2100ºC, are typical formaterials that undergo dynamic recrystallisation.After a very steep rise in the stress-strain curve at

Fig. 8 SEM imagesof the surface areanear the fracture forpure iridium afterstress-rupture test at1800ºC under aninitial load of 18MPa: (a) ×10; (b)×50; (c) ×500; (d)×200

Fig. 9 Stress-straindiagram for pureiridium at differenttemperatures


the beginning of the tests, the slope of the curvedecreases, probably due to both plastic deforma-tion and softening of the material caused bydynamic recovery. When dynamic recrystallisationcommences, the softening becomes more severeand the load decreases. This is followed by a peri-od at a more or less constant load – in some casesthis phase exhibits a slight oscillation – until thestress drops rapidly when the sample ruptures.This behaviour is typical for materials that arerecrystallising dynamically during the high-temper-ature tensile tests.

A comparison of the stress-strain curves withthe results of the SEM investigations reported inthe previous section shows that the slip bands,shown in Figures 8 and 10, are formed only in sam-ples exposed to loads close to or greater than theyield stress determined in the tensile test at the cor-responding temperature. In the presentinvestigation, the deformation mechanism underhigh loads at high temperatures appears to bebased on an interaction of plastic deformation dueto dislocation slip with common creep deforma-tion. The plastic deformation in turn leads to anincrease in the internal deformation energy, thuspromoting the initiation of dynamic recrystallisa-tion.

Yield strength, Rp0.2, tensile strength, Rm, andtensile elongation, A, as determined in the high-temperature tensile tests are plotted in Figure 11 asa function of temperature. Measurements were

made on at least three samples for each test tem-perature. The values given for Rp0.2 and Rm areaverages over all three measurements. The stan-dard deviation is indicated as an error bar for eachdata point, revealing excellent reproducibility forthe measurements. The values for A, however,showed a greater scatter. Moreover, some of thesamples did not rupture between the markers, inwhich cases it was not possible to determine thetensile elongation with the “SuperCreep” software.No error bars are given for the relevant data pointsin Figure 11.

Metallographic InvestigationsThe microstructure of pure iridium exposed to

the influence of high temperatures and different

Fig. 11 Yieldstrength, Rp0.2, ten-sile strength, Rm,and tensile elonga-tion, A, of pureiridium as a functionof temperature

Fig. 10 SEM image (×200) of the surface morphology ofan iridium sample after creep test at 1800ºC under 23MPa load


loads during the creep tests was evaluated metallo-graphically. Comparative investigations werecarried out on longitudinal sections of samplesbefore and after creep testing. The microstructureof a sample in the initial state (i.e. before creeptesting) is shown in Figure 12. The sample exhibitsa uniform microstructure with an average grainsize of about 100 μm.

Comparison with metallographic sections ofsamples that were exposed to high temperaturesunder different loads (Figures 13 and 14) showsthat during creep tests specimens have undergonethe expected severe grain coarsening. Due to thehigh temperatures and long test times, very coarsegrains with grain sizes up to 4 mm have formed.Furthermore, the samples exhibit strong intercrys-talline crack formation as mentioned above.Figures 13 and 14 show that several – partly verydeep – cracks have formed within the same sam-ple. Nevertheless, the material withstood this

damage and cracked at a different position, sever-al hours later. Dynamic recrystallisation isapparent in areas of high stress concentration andstrong deformation, for instance at crack tips(Figure 13(a)), and close to the fracture in thenecking areas (Figure 15). Particularly in the areaaround the crack tips, this may have led to a reduc-tion in local stresses. As a consequence the crackpropagation may have stopped in this area andcontinued in another part of the sample.

The metallographic images show the existenceof individual voids along grain boundaries andtheir coalescence into large pores. These are typi-cally found in materials that have undergone creepdeformation. This phenomenon is most frequent-ly observed on grain boundaries that are orientedperpendicular to the direction of applied stress.This often leads to the formation of intercrys-talline creep cracks during the final period of creepdeformation.

Fig. 12 Longitudinal section of pure iridium in the initial recrystallised state

Fig. 13 Longitudinal sections of pure iridium after creep test at 1800ºC, 6.7 MPa, 1403.7 h


Microanalytical InvestigationsAs has been demonstrated in previous investi-

gations (14–19), iridium generally tends to brittleintercrystalline fracture due to trace impurities atgrain boundaries. Investigation by SIMS on thepresent iridium samples has shown a very highpurity for the material. Only very small traces ofimpurities in the ppm range were detected, show-ing no enrichment at the grain boundaries. All theelements detected were distributed homogeneous-ly in the matrix.

ConclusionsDue to its outstanding properties iridium is par-

ticularly suited to applications under extremethermal, chemical and mechanical conditions. Inorder to obtain the materials data necessary for thedesign of high-temperature equipment and the

numerical simulation of its service performance,the stress-rupture strength, creep behaviour andtensile properties of iridium have been measuredover the temperature range 1650–2300ºC.

The investigations were performed on hotrolled iridium sheet. The results showed a verygood degree of reproducibility. The iridium sam-ples exhibit very high stress-rupture strength. Adiscontinuity in the slope of the stress-rupturecurves indicates the existence of two different frac-ture modes, depending on the temperature and theinitial load applied to the samples. The existence ofdifferent fracture mechanisms was confirmed bythe examination of the fracture surfaces. Thechange in fracture mode is probably caused by dif-ferent deformation mechanisms prevalent underthe various test conditions.

A significant anomaly was observed in the creepbehaviour of pure iridium – the creep curves con-tained plateaus in the range of steady-state creep.Metallographic examination, investigations bySEM, and high-temperature tensile tests indicatedthat dynamic recrystallisation may be the cause ofthis phenomenon. A further influence may be theactivation of various slip systems which can bededuced from the observed slip bands.

Microanalytical investigations by means ofscanning SIMS showed a very high purity of theiridium heats investigated, without any enrichmentof trace impurities at the grain boundaries.

Initial results on the stress-rupture strength ofan iridium-rhenium alloy doped with hafnium andmolybdenum indicate that this alloy exhibits rup-

Fig. 14 Longitudinal sections of pure iridium after creep test at 1800ºC, 13 MPa, 56.8 h

Fig. 15 Longitudinal section through the fracture tip ofpure iridium after creep test at 1800ºC, 8.3 MPa, 385.9 h

The AuthorsAfter studying MaterialsScience at the Universityof Saarbrücken, Germany,Reinhold Weiland workedas a research associate atthe Max-Planck-Institute

for Metals Research in Stuttgart, andreceived his doctoral degree from theUniversity of Stuttgart. Since 2002 he hasbeen working at W.C. Heraeus GmbH inHanau as a development project manager.His main interests are the processing andmanufacture of precious metals productsand composite materials.

Carolin Scheckenbachstudied materialstechnology at theUniversity of AppliedSciences in Jena,Germany, where she

received her diploma degree. After oneyear as a research associate, she is nowworking as a trainee in the Department ofMelting Technology and Hot Forming atSchott AG, Mainz.

Jürgen Merker studiedmaterials science at theTechnical University,Dresden, where he alsoearned his Ph.D. degree.Between 1995 and 2000,

he worked as a development projectmanager for W. C. Heraeus GmbH, Hanau.After a short time in the R&D departmentof KM Europa Metal AG, Osnabrück, hewas appointed to a Professorship inMaterials Engineering and Applied MetalsScience at the University of AppliedSciences Giessen-Friedberg in 2002. In2006 Dr Merker became Professor ofMaterials Technology and Materials Testingat the University of Applied Sciences Jena.

Prof. Dr David Lupton isDevelopment Manager,Engineered MaterialsDivision, W. C. Heraeus.His main interests are themanufacture and applic-

ations of precious metal and refractorymetal products.

Prof. Dr.-Ing. habil. BerndFischer studiedmechanical engineeringand materials science atthe Technical UniversityChemnitz, Germany. After

more than 25 years at the University ofJena, Bernd Fischer was appointed to theChair of Materials Science at the Universityof Applied Sciences Jena in 1992. Formany years, his research interests haveincluded the properties and applications ofnoble and refractory metals.

Jörg Witte studiedchemistry at the TechnicalUniversity of Darmstadt,Germany, from which hereceived his doctoraldegree in Materials

Science. Since 1999 he has worked atSchott-AG, Mainz, where he is currentlyhead of the Materials CompetenceDepartment. His main interests are theproperties of refractory materials andfailure analysis.

ture times three to four times longer than for pureiridium. Moreover, this alloy shows outstandingductility, comparable to that of pure iridium. Thisalloy is therefore of particular interest for high-temperature applications and is the subject ofongoing research.

AcknowledgementsThe authors would like to thank Margit

Friedrich (SEM investigations), Erik Hartmann andFrank Lehner (creep tests) from the DepartmentSciTec of the University of Applied Sciences Jenafor their support for these investigations.

References1 “Edelmetall-Taschenbuch”, eds. G. Beck, H.-H.

Beyer, W. Gerhartz, J. Hausselt and U. Zimmer,OMG AG & Co KG, Giesel Verlag GmbH,Isernhagen, 2001

2 C. R. Barber, Platinum Metals Rev., 1969, 13, (2), 65 3 B. Fischer, A. Behrends, D. Freund, D. F. Lupton

and J. Merker, Platinum Metals Rev., 1999, 43, (1), 184 J. Merker, D. Lupton, M. Toepfer and H. Knake,

Platinum Metals Rev., 2001, 45, (2), 745 D. F. Lupton and B. Fischer, Proceedings of the

Second European Precious Metals Conference,Lisbon, 10th–12th May, 1995, Eurometeaux,Brussels

6 J. Merker, B. Fischer, D. F. Lupton, C.Scheckenbach, R. Weiland and J. Witte, Proceedingsof Processing and Fabrication of Advanced

Materials XIII, Singapore, 6th–8th December, 2004,Stallion Press, Singapore, 2005, pp. 787–799

7 R. Voelkl, D. Freund and B. Fischer, J. Test. Eval.,2003, 31, (1), 35

8 R. Voelkl and B. Fischer, Exp. Mech., 2004, 44, (2),121

9 J. N. Greenwood and H. K. Worner, J. Inst. Met.,1939, 64, 135

10 M. J. Luton and C. M. Sellars, Acta Metall., 1969, 17,(8), 1033

11 R. E. Reed-Hill and R. Abbaschian, “PhysicalMetallurgy Principles”, 3rd Edn., PWS PublishingCompany, Boston, 1994

12 C. Herring, J. Appl. Phys., 1950, 21, (5), 43713 F. R. N. Nabarro, Proceedings of the Bristol

Conference on Strength of Solids, Physical Societyof London, 1948, p. 75

14 P. Panfilov, A. Yermakov, V. Dmitriev and N.Timofeev, Platinum Metals Rev., 1991, 35, (4), 196

15 P. Panfilov and A. Yermakov, Platinum Metals Rev.,2001, 45, (4), 179

16 J. Merker, M. Schlaubitz, H.-J. Ullrich, S. Garbe, A.Knoechel, M. Radtke, F. Lechtenberg, D. F. Luptonand B. Fischer, DESY-HASYLAB Annual Report,Hamburg, Germany, 1994, pp. 913–914

17 J. Merker, D. F. Lupton, B. Fischer, M. Schlaubitz,H.-J. Ulrich, A. Gebhardt and S. Garbe, Prakt.Metallogr., Sonderband 27, 1995, pp. 267–270

18 J. Merker, D. F. Lupton, H.-J. Ullrich, M. Schlaubitzand B. Fischer, Proceedings of the TMS AnnualMeeting, 2000, TMS, Warrendale, Pennsylvania, pp.109–120

19 R. Voelkl, A. Behrends, J. Merker, D. F. Lupton andB. Fischer, Mater. Sci. Eng. A, 2004, 368, (1–2), 109


doi: 10.1595/147106706x154198 high-temperature mechanical ... · fig. 1 schematic diagram of...

Documents