The Distribution of Boron in Pure IronA. Brown, J. D. Garnish, and R. W. K. Honeycombe

The solubility of boron in pure iron has been deter-mined over the temperature range 500-1300 C usingthe technique of boron autoradiography, which candistinguish between boron in solution, boron-richprecipitates, and boron segregated to grain boundaries.However, the solubility limits determined should beregarded as upper limits, as the technique cannotdifferentiate between soluble boron and very fine borideparticles 600A dia.). In y-iron, the boron segregatesuniformly to the grain boundaries, the degree of segre-gation reducing with increasing temperature. However,in cx-ironprecipitation of iron boride at grain boundarieswas observed, with no detectable segregation betweenthe particles. A Snoek peak was not detected in internal-friction measurements, indicating that boron is not ininterstitial solid solution in cx-iron.The segregation toaustenite boundaries delays the nucleation of ferriteleading more readily to an acicular ferrite morphology.The structures obtained on quench-ageing were alsoexamined.

The improvement, in the mechanical properties of iron andsteel brought about by trace quantities of boron has beenrecognized commercially for over 20 years.l,2 Properties wellknown to be affected by boron include hardenability offerritic steels,3 hot-working characteristics of austeniticsteels,4 and creep strength of both austenitic and ferriticsteels.5 In all these instances, boron exerts its most significanteffects at levels < 120 ppm, and indeed above this level someproperties, e.g. ductility, deteriorate again.The mechanism by which boron' influences these properties

is in many cases not clearly understood, but it seems certainthat the predominant effects are interaction with latticedefects and grain boundaries.6 In alloys containing carbon ornitrogen there is also an effect on precipitation' behaviourwhich in' many cases may be ascribed to the defect-seekingnature of the boron atom.7 In view of the extremely low con-centrations of boron employed, an understa.pding of thesolubility of boron in these materials is' essential to the elu-cidation of the mechanism of these effects. A survey of theliteratureS,17 shows that while a number of workers haveinvestigated the nature of the solution of boron in iron,rather fewer attempts have been made to determine the

Paper No. MS 374. Manuscript received 22 Febru';lry 197~ ..A. Brown,PhD, is now with the Science Staff Ext~rnal SerVIces, BntIsh Broad-casting Corpora~ion bl;lt was fo!mer.1Y III the De'part~ent of Me~al-lurgy and Matenals SCIence, Umversity of Ca~bndge, J. D.-Garmsh,PhD is in the Division of Applied ChemIstry, AERE, Harwell;Prof~ssor R. W. K. Honeycombe, DSc, PhD, FIM,. is in the De.part-ment of Metallurgy and Materials Science, UniverSIty of Cambndge.

METAL SCIENCE317

solubility. It is apparent that boron interacts strongly withother elements, particularly interstitials, which may bepresent in the iron, and indeed the values of solubility re-ported by various workers have tended to lower levels as thepurity of the materials increased. 9The nature of the solution of boron in 'cx-iron is still in

dispute,10-12 some experiments suggesting an interstitialsolubility, while others indicate substitutional behaviour. Auseful summary of the evidence has been presented by Lucciand Venturello.ll There is some evidence to suggest that boronforms a true substitutional solid solution in iron at temper-atures lower than 900 C, but an alternative explanation is thatboron in a substitutional location is stabilized by an adjacentinterstitial impurity atom. 8,9 This would explain the conflic-ting internal-friction data and the variation of boron solu-bility with purity. There is less evidence about the nature ofthe solution' in y-iron, S but it is probable that the samen1echanism will apply. At all temperatures, there will also bea contribution from boron bound to lattice defects.. For these reasons, it was decided (a) to re-determine thesolubility curve for boron in high-purity iron at temperaturesup to 1250 C, using for the analysis of boron the autoradio-graphic technique,15,16 which can eliminate the contributiondue to segregation effects ; (b) to examine the microstructurein quench-aged specimens using high-resolution autoradio-graphy and electron microscopy; (c) to re-investigate thenature of the solution in cx-iron using internal friction andstrain-ageing techniques.

Specimen PreparationThe starting material was Glidden AI04B high purity

electrolytic iron flake) which, is low in substitutional andinterstitial impurities except for a high oxygen content. Theanalysis of this starting material is shown in Table 1. Flakewas consolidated into a 750 g bar in a horizontal 'zone re-finer using" a cold-crucible technique13 and a high-purityargon/hydrogen atmosphere. The bar, which was'" 600 mmlong with an oval cross-section of '" 18 X 12 mm, was given42 zon-e-refiningpasses. It was then cut into sections, polished,and treated with hydrogen at 1480 C for -48 h in a high-purity alumina tube. The analysis* of this iron after 27 passesplus the hydrogen-treatment is shown in Table I, from whichit can be seen 'that the total non-metallic concentration was22 ppm, while Battelle iron analysed* in the same way.had18 ppm. Boron alloys (25-100 ppm)t were made up by dIrectmelting of this material with spectrographically pure zone-refined boron (99'995%) in the water-cooled copper boat

*We are very much indebted to Dr. E. J. McLauchlan of the NationalPhysical Laboratory for carrying out thes~ analyses. .tThroughout this paper, boron concentratIOns are expressed III parts permillion by'weight.

1974. Vol 8

Solubility DeterminationAfter preliminary experiments to establish the approximate

range of boron solubilities and heat-treatments, a series ofspecimens was prepared (see Table II). All specimens were

.under a high-purity argon atmosphere. The alloys werehomogenized by giving several zone-levelling passes in thehorizontal melter followed by treatment for 24 h at 1150Cin a silica capsule evacuat~d to 10-6 torr (013 mN/m2). Theboron contents of the alloys were subsequently confirmedusing autoradiography.

Heat-Treatment and QuenchingIn the preliminary experiments, specimens were solution-

treatedfor 40 h at 1150C in silica capsules and then quen-ched in water without breaking the capsule. This avoidedcontamination and deformation at this stage.However, it was found that specimens quenched in this

manner from the y-region showed' evidence of borideprecipitation at the sub-grain boundaries. Since these boun-daries are formed during cooling, it was clear that a morerapid quench was necessary to retain boron in solution fromhigh temperatures. This quench-rate sensitivity has also beenreported by Goldschmidt.14In order to achieve higher quench rates, specimens were

prepared in the form of plates '" 5 X 1Ox 06 mm. Thesewere heat-treated in a 1 MHz induction furnace fitted with atantalum susceptor. Specimens were supported from a thinwire of the same composition and temperatures were meas-ured by means of shielded thermocouples spot-welded to thespecimens. A dynamic vacuum of 10-6 torr (013 mN/m2)was maintained during heat-treatment and specimens werequenched in situ by a high-velocity stream of He/5% H2.Quenching rates of up to 1000CIs were measured; no furtherevidence was obtained of boron mobility during this type ofquenching. .

TABLE I

Analyses of Pure Iron

IUnprocessed 27 pass iron +

Element Glidden AI04B . hydrogen-treatn1entppm ppm

Al n.d.* n.d.Bi n.d. n.d.Co 15 15Cr n.d. n.d.Cu 10 2Mn n.d. n.d.Mo n.d. n.d.Ni .15 15Ph n.d. n.d.Si n.d. n.d.Sn n.d. n.d.Ti n.d. n.d.B n.d. n.d.V n.d. n.d.0 200 7H I

Brown, Garnish, and Honeycombe: The Distribution of Boron in Pure Iron 319

/o /(1I /I /

~: // y+-1iq.~

a

Fig. 2 The solubility of boron in oc- and y-iron.

1000

1100

1200

~900 //~-W ,/ ...-0:: a. / ,,/::> ~_/~ //WO::SOO ~/

I I

0... //2 Pw 'II- 700 ,

II

Figs. lea) and (b) Typical boron alltoradiographs,o A boron"ri~h pre-cipitates,o B boron segregated to grain boundaries; C jlnron insolution.

boundaries, and in these regions a uniform distribution ofisolated tracks is obtained. It is not possible, however, toascribe all these tracks unambiguously to 'soluble' botonatoms for the foIIowing reason. If the region of interestcontains, for example, 1 ppm, then, in order to obtain therequired track density,""" 1 boron atom in 2 X 106 mustundergo fission._ In a large precipitate, therefore, severalatoms react and a 'star' of tracks is formed in the film. As theprecipitate size decreases, however, fewer and fewer boronatoms undergo fission in the particle and a situation is reachedwhere only one or two tracks are produced by a given par-ticle. Under the conditions described, an Fe2B precipitate of,....,600A dia. would produce, on the average, only two tracks.Hence, these tracks would not be recognizable in the auto-radiograph as originating from a precipitate and would becounted as 'soluble' boron. For a given track density, theproportion of boron atoms undergoing fission in anyoneexperiment is inversely proportional to the boron content,so that the critical volume of an 'undetectable' precipitatewiII increase with the apparent 'soluble' boron content. Thus,if the neutron dose is adjusted to give a track density of 107/cm2 from a specimen with a uniformly distributed boroncontent of 70 ppm, the smaIIest detectable Fe2B precipitatewiII be rv 2500A dia. For these reasons, the best results willbe obtained from samples with a high supersaturation ofboron, which will tend to produce large precipitates, and theleast reliable results from samples where the boron contentonly slightly exceeds the solubility. Summarizing, it is not

normaIIy possible to distinguish between 'soluble' boronatoms and boron atoms segregated to dislocations or to pre-cipitate particles of less than a certain critical volume, andall these will be included in the solubility figure so obtained.In any event, the boron concentration observed by auto-radiography will clearly represent an upper limit for the truelattice solubility.For analysis, the prepared samples were nlounted in

Bakelite,' ground to one-half thickness to eliminate surfaceeffects, polished, and examined autoradiographicaIIy by thestandard published procedure.15,16 Preliminary experimentshad indicated that the maximum solubility of boron in iron,which occurs at the eutectic temperature (I'V 1150 C), doesnot exceed 70 ppm and the samples examined had overallboron contents of 50-100 ppm. A series of high-resolutionautoradiographs was obtained for a variety of thermalneutron doses to give a rough estimate of the boron in solutionin each sample. Using these results, each sample was auto-radiographed again with a neutron dose calculated to pro-duce 107 tracks/cm2 over the precipitate-free regions. Thethermal neutron dose was monitored using calibrated goldfoils, and- the film background determined over a high-puritysilicon substrate.16 After developing each autoradiograph inthe normal way, it was examined under an optical phase-contrast microscope and the track density (less background)determined over a number of different grains. To avoidconcentration effects, no counts were taken within 10 (Lmofa visible precipitate or a grain boundary. In no case weresignificant .variations in track density recorded across asample. From the track density so obtained, and knowing theneutron dose received by each sample, the boron 'in solution'

320 Brown, Garnish, and Honeycombe: The Distribution of Boron in Pure Iron

at each temperature was calculated. The results are shown inFig. 2, ,together with the estimated limits of error. It will benoted that the maximum error occurs at the higher solubilities,a feature arising from the low supersaturation mentionedabove. In every sample examined, saturation had been a-chieved and precipitates were present but these were only justdetectable in the samples aged above 1000 C.For this reason,rather large uncertainty values have been placed on the re-sults obtained from these samples but there is every reason tobelieve that the correct figure lies well towards the low end ofthe spread.Typical autoradiographs taken from this range of samples

are shown in Fig. 3(a)-(f). For this series of autoradiographsthe neutron doses were chosen to reveal the metallographicfeatures of the samples and were not necessarily those usedfor quantitative analysis.

Microstructure of Quench-Aged SpecimensElectron microscopy was used to examine the early stages

of the boride precipitation in ferrite after quenching from anaustenitic solution-treatment. Taking into account the solu-bility at the eutectic temperature as determined by the auto-radiographic analysis. previously described, specimens con-taining ,....,50 ppm boron were given a quench-ageing cyclein the vacuum-quenching furnace. The cycle was solution-treatment at 1140 C for 15 min and quenching to roomtemperature, followed by ageing at temperatures between200 and 750 C for times ranging from 1 min to 3 days, thenquenching to room temperature.Some initial difficulty was experienced in preparing thin

foils and surface replicas using standard techniques whenparticles of the boride Fe2B were present in the microstruc-ture. This was attributed to selective attack of the borideduring polishing. Optical examination of surfaces polishedor etched, using conventional chemical or electropolishing-solutions, support this conclusion, as etch-pit patterns aredetected corresponding to the boride distribution observedautoradiographically. Satisfactory specimens were obtained

(a)

(b)

(c)

Brown, Garnish, and Honeycombe: The Distribution of Boron in Pure Iron 321

(d)

(e)

(f)

Fig. 3 Distribution of boron in (X- and y-iron. Boron autoradiographs: (a) y-iron 1250C; (b) y-iron 1100 C,'(e) y-iron 950 C; (d) (X-iron 900 C; (e) (X-iron 700 C; (f) (X-iron 500 C.

ultimately with potentiostat control of the polishing con-, dition.I7

No precipitation was observed in specimens as quenchedafter solution-treatment. On ageing, the initial distribution ofboride was seen most clearly after 30 h at 450 C. In thiscondition, particles of t'V 600A in size have formed in the(X sub-grain boundaries Fig. 4. However, not all boundariesare decorated with precipitation; on the average about oneboundary in three is found in this condition. At 500 C, asimilar distribution of slightly larger particles is detected andthe spacing between boundaries with precipitation agreeswith the autoradiographs for that temperature. The volumefraction of the precipitate is too low to give well-definedelectron-diffraction patterns, but X-ray-diffraction studies

gave a. tentative identification of the particles as the equi-librium precipitate Fe2B.The above observations related to the microstructures

where the quenching rate following solution~treatment wasestimated to be t'V 500 Cis. Using slightly thinner foils andtHemaximum rate possible in the vacuum-quenching furnace(t'V 1000 Cis), a radically different structure was observed(Fig. 5). This consists of long, thin laths t'V 05 flmwide witha high dislocation density and no internal twinning, verysimilar to that found in low-carbon martensite. However,electron-diffraction patterns showed that this phase had abcc structure. It is therefore postulated that the structure isacicular ferrite, formed from austenite by a shear transfor-mation as a result of the segregation of boron to the austenite

322 Brown, Ga..rnish,and Honeycombe: The Distribution of Boron in Pure Iron

Fig. 4 Iron boride in subgrain boundaries of an Fe-50ppm B alloysolution-treated 15 min at 1140 C, quenched, then aged 48 hat 400 C. Electron micrograph. X 50000.

Fig.5 Acicular ferrite in an Fe-lOOppm B alloy, solution-treated15 min at 1140 - C, then quenched at 1000 Cis. Electronmicrograph. X 26 500.

I

25120

A

I

20

B

90

ageing temperature 400C

I

5I

oo 30 60

,TIME, min10 '15

TIME2/3 min2/3,Fig. 7 Strain-ageing curves: A pure iron,. B Fe-l 00 ppm B.

~~ 01w~z0:::l-ll)

and Wells19 suggest that the mobility of the boron atom isabout the same as the self-diffusion of iron at low temperatures.Accordingly, the temperature range for ageing was extended.up to 5000 C.

Modified Hounsfield strip-tensile specimens were annealedat 9000 C and quenched. On straining 7'5%, all specimensdisplayed a yield drop and a lower yield-point extension.However, the iron-boron alloys showed some reduction inthese effects compared to the pure irons. Yield-point pheno-mena in very pure iron have previously been shown24,25 to bea result of the low oxygen level produced by dry-hydrogentreatment. These effects are not detected with wet-hydrogentreatment owing to the formation of C-O couples.

On ageing, it was observed that strain-ageing diminished asthe boron content increased Fig. 7, while the behaviour ofthe pure iron reflects the residual carbon and nitrogen in thematerial; the decrease of strain-ageing on the addition ofboron suggests that this element is scavenging interstitialsin solution from the iron. In the ageing curves determinedfor the various iron-boron alloys, no features were found thatcould be attributed to the segregation of boron to dislocationsover the range 20-500 C.

DiscussionThree groups of workers have published extensive measure-

-:02

Brown, Garnish, and Honeycombe: The 'Distribution of'Boron in Pure Iron 323

from the photographs presented here (which represent only'" 10-3-10--4 cm2). It appears that there is a marked differencein the nature of the segregation of boron to the grain boun-daries in rx- and y-iron. In those specimens quenched from they region, no trace of precipitation is detectable in the grainboundaries by autoradiography, and the segregation appearsto be continuous. As expected,23 the degree of segregationdiminishes with increasing temperature. In contrast, all th'especimens quenched from the rx region reveal precipitation atthe grain boundaries, with no detectable continuous segrega-ti0l! between the precipitates. The difference is illustrated inthe autoradiographs of the two specimens. quenched from950 C (y) and 900C (rx) (Figs. 3 {c) and (d).These specimenscontain the same total boron concentration (80 ppm) and thesame measured boron content in solution (20 ppm).The absence of a Snoek peak for boron in the internal-

friction spectra provides strong evidence to suggest that theelement does not form an interstitial solution in rx-iron. Thisargument is reinforced as a result of the increased carbon~elaxation observed in the presence of boron. The formationof substitutional boron-interstitial carbon pairs appears tobe the most likely mechanism to explain this effect. 8,9 Thistype of interaction may also be the explanation for thesuppression of strain-ageing behaviour in the presence ofboron. In this case, it can be argued that the formation of theB-C couples restricts the long-distance mobility of the carbonatom. However, as a residual level of nitrogen is present inthe iron used in this investigation, the mechanism may alsoinvolve the removal of nitrogen from solution by the for-mation of boron nitride, as discussed by Butler. 26Although boron is not an interstitial solute in rx-iron, the

pattern of precipitation observed for the boride Fe2B on aquench-ageing treatment is clearly very similar to that pre-viously reported for the high-temperature precipitates Fe3Cand Fe4N in the Fe-C and Fe:-N systems.27 However, asmight be expected, the precipitation of boride is very muchslower, perhaps reflecting the much higher activation energyfor diffusion of boron in rx-iron.19The formation of acicular ferrite at high quenching rates

can be interpreted as a manifestation of the same action ofboron as in the well-known hardenability effect.2,3 In thepure Fe-B system, the segregation of boron to the normalnucleation sites for polygonal ferrite may delay the nucleationand growth of this phase to lower temperatures where theshear reaction is favoured.As discussed earlier in this paper, the solubility limits

measured for boron in rx- and y-iron in the present investiga-tion have to be regarded as upper limits. The figures do notdistinguish between soluble boron atoms and boron segre-gated to the dislocations, but this inaccuracy is not likely toexceed 1 ppm. Owing to the likely formation of the B-Ccouples mentioned above, a far larger inaccuracy arises fromthe presence of the residual levels of carbon and nitrogen inthe iron. Taking the minimum value from the uncertaintyspreads for rx-iron, it is possible to argue9 that the total solu-bility measured in this phase could be due to the stabilizationof boron by interstitial impurities. Even if this is not completelytrue, it seems likely that lower solubility limits would bemeasured in iron of even greater purity. For y-iron, on theother hand, even after all restrictions have been applied, wecan be certain that thermodynamically normal lattice solutionof boron occurs at temperatures above 950 C, but, on theevidence available from the present investigation, no con-clusion can be reached as to whether this is an interstitial orsubstitutional solid solution.

r-------fr-t

max. 210ppmat 1170C

~

this'work

Nicholson

:tf ..Ii:/

'7

max. 210ppmat 1149C

Busby et al.

McBride et al.

11

12

7

w....60:::Jt-12

~10I-

6o 20 40 60 80 100 20 40 60 80 100

BORON CONTENT, ppm by weight

Fig. 8 Phase diagrams of the Fe-B system (low-B end).

ments. on the low-boron end of the iron-boron system,20-22and portions of the phase diagrams constructed from theirdata are shown in Fig. 8. Other workers have publishedisolated solubility data and where the quoted values lie below250 ppm these are also shown. Figures higher than this,common among the earlier publications, are almost certainlyin error and have been ignored.

. The values for solubility obtained in the present work agreeclosely with those obtained by Nicholson21 at temperaturesbelow 950 C, though they are significantly lower at highertemperatures. As discussed above, the present work can de-termine only an upper limit for the amount of boron held insolution in the iron lattice. and it is not possible to deter-mine the proportion of this boron bound to defects. However,the figure obtained does represent the limiting amount ofboron that can be held at equilibrium within the metal grainswithout giving rise to detectable precipitation and ignores theboron segregated to grain boundaries. In this work, thegrain size of those specimens quenched from the y-region islarge (see Fig. 3) and the fraction of boron segregated to theboundaries is insignificant. The fact that the solubility figuresobtained in this way are. still lower than the values publishedpreviously may be attributed to the lower impurity content,particularly of interstitials, in the material used for this work.In order to. complete this portion of the phase diagram, it is

necessary to establish the shape of the curve around therx-y transformation temperature. Lucci et al.9 showed bymeans of dilatometry that the minimum at '" 10 ppm boron,which was postulated for the peritectoid reaction by McBrideet af.,22 does not exist, and that the equilibrium region is verynarrow. This conclusion is supported by dilatometric studiescarried out by the present authors. '

One further observation from the autoradiographic studyarose from the analysis of all the autoradiographs (a total of'" 20 cm2 of metal surface) but cannot be accurately deduced

324 Brown, Garnish, and Honeycombe.o The Distribution of Boron in Pure IronAcknowledgements

We gratefully acknowledge support of the work by theBritish Steel Corporation and the Science Research Council.

References1. T. G. Diggs, C. R. Irish, and N. L. Carwile, J. Res. NBS, 1948,

41,545.2. R. A. Grange and T. M. Garvey, Trans. Amer. Soc. Metals,

1946,37, 136.3. R. A. Grange and J. B. Mitchell, ibid., 1954, 46, 446.4. F. E. White, Laminoirs, D.P.H.T., 1967, 152, 39.5. F. C. Hulland R. Stickler, 'Froceedings of aJointlnternational

Conference on Creep', pA9, 1963: Lonuon(Inst. Mech. Eng.).6. G. Henry and J. Philibert, Mem. Sci. Rev. Met., 1970, 67, 233.7. B.M. Kapadia, R. M. Brown, and W. J. Murphy, Trans. Met.

Soc. AIME, 1968, 242, 1689. .8. J. D. Garnish, AERE-R 6155, 1969.9. A. Lucci, G. Della Gatta, and G. Venturello, Metal Sci. J.,

1969,3, 14.10. Y. Hayashi and T. Sugeno, Acta. Met., 1970, 18, 693.11. A. Lucci and G. Venturello, Scr. Met., 1971, 5, 17.12. Y. Hayashi and T. Sugeno, ibid., p. 25. -

13. J. Beech, B. A. Hands, and P. P. Mohla, Lag. Methods, 1967,76,175.

14. H. J. Goldschmidt, J. Iron Steel Inst., 1971. 209, 900.15. J. D. H. Hughes and G. T. Rogers, J. Inst. Metals, 1967, 95,

299.16. J. D. Garnish and J. D. H. Hughes, J. Mat. Sci., 1972,7, 7.17. A.-Brown, PhD Thesis, Univ. Cambridge, 1973.18. M. E. Nicholson, Wright Air Development Center, Tech. Dept.

5, 1955.19. P. E. Busby and C. Wdls, Trans. Met. Soc. AIME, 1954, 200,

972.20. P. E. Busby, M. E. Warg:l, anj C. Wdls, ibid., 1953,197, 1463.21. M. E. Nicholson, ibid., 1954,200, 185.22. C. C. McBride, J. W. Spretnak, and R. Speiser, Trans. Amer.

Soc. Metals, 1954,46,499.23. W. F. Jandeska, PhD Thesis, Univ. Illinois, 1971.24. B. F. Oliver and F. Garofalo, Trans. Met. Soc. AIME, 1965,

233, 1318. .25. P. R. Mould and G. V. Smith, Mem. Sci. Rev. Mh., 1968,

55,271.26. J. F. Butler, Trans. Met. Soc. AIME, 1962, 224, 89.27. A. S. Keh and W. C. Leslie, Mat. Sci. Res., 1963, Vol.

(edited by H. H. Stadelmaier and W. W. Austin).

The Metals Society. 1974.

Conference on

THE MECHANICS AND PHYSICS OF FRACTURE

A Conference on The Mechanics and Physics of Fracture is being arranged by the Materialsand Testing Group of The Institute of Physics in collaboration with The Metals Society, totake place at Churchill College, Cambridge, on 6-8 January 1975.

-l

The programme contains invited papers reviewing recent developments and cl:Jrrent workunderJour main headings, which will be covered in one morning or afternoon session each.

The Mechanics of FractureChairman: professor J. D. Eshelby

II Fracture and MicrostructureChairman: Professor R. W. K. Honeycombe

III Intercrystalline FractureChairman: Professor G. W. Greenwood

IV FatigueChairman: Dr. F. J.rBradshaw

Guest of Honour at the Conference d!nner will be SirAlan Cottrell, FRS

Proceedings will not be published, but a handbook containing the papers will be producedin advance of the Conference and will be available afterwards on general sale.

Further details are available from the M.eetings Office, The Institute of Physics, 47 BelgraveSquare, London, SW1 X 80X. . - ,