Lead and lead alloy foams - MITweb.mit.edu/3.042/team3_11s/papers/irretier2005.pdfLead and lead...

Acta Materialia 53 (2005) 4903–4917

www.actamat-journals.com

Lead and lead alloy foams

Andree Irretier a,1, John Banhart b,*,1

a Institut fur Werkstofftechnik, 28359 Bremen, Germanyb Hahn-Meitner-Institut and Technical University Berlin, 14109 Berlin, Germany

Received 25 April 2005; received in revised form 4 July 2005; accepted 7 July 2005Available online 26 August 2005

Abstract

Lead-based foams were produced by applying the powder route, i.e., by mixing metal powders and a powdered gas-releasingblowing agent and pressing the mixture to a foamable precursor material afterwards. This precursor was then foamed by heatingit up to above its melting point, thus triggering gas release and foam formation. In a first step a set of process parameters was estab-lished, allowing us to foam lead alloys. This included finding suitable metal and blowing agent powders, mixing and compactionprocedures, and determining appropriate foaming conditions. After this the foaming process of various alloys was studied as a func-tion of time, temperature, alloy composition and blowing agent content.� 2005 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Powder processing; Foaming; Lead; Foams

1. Introduction

Cellular metals can be produced by a variety ofdifferent methods [1]. A versatile and promising methodfor producing closed-cell metal foams from metalpowders was rediscovered in the early 1990s [2] after ithad been developed long ago [3]. The process comprisesmixing metal powders and a powdered blowing agentand compacting the mix to a dense semi-finished product(called ‘‘foamable precursor material’’) by hot pressing,extrusion, powder rolling or other methods. In the actualfoaming step the foamable precursormaterial is expandedby heating it up to above its melting point. This transfersthe metal into a semi-liquid viscous state and simulta-neously makes the blowing agent decompose, thus releas-ing gas and creating a highly porous structure [4].

Up to now the main focus in the literature has beenon aluminium and aluminium alloy foams [5,6]. The

1359-6454/$30.00 � 2005 Acta Materialia Inc. Published by Elsevier Ltd. A

doi:10.1016/j.actamat.2005.07.007

* Corresponding author. Tel.: +49 30 8062 3059.E-mail address: [email protected] (J. Banhart).

1 Formerly at Fraunhofer-Institute for Advanced Materials, 28359Bremen, Germany.

present paper deals with lead and lead alloy foams.There is some technological interest in lead foams be-cause of possible applications: e.g., they could be usedas light-weight electrodes in lead-acid batteries. As leadfoams are superconducting below 7 K, their use as cryo-genic shielding material has been proposed [7]. Further-more, lead foams are interesting because they allow usto study foaming processes in metals fairly easily atlow temperatures. As lead has a high density of11.34 g/cm3, such foams are being used for comparativestudies of foams under normal and zero gravity condi-tions, e.g., for studying drainage phenomena [8].

To be able to produce high-quality lead foams theprocess parameters have to be chosen carefully. This in-cludes an appropriate selection of the metal powders,the alloy composition, the type of blowing agent, com-paction conditions and, finally, foaming temperatures.Establishing these parameters was one purpose of thepresent work. We succeeded in creating uniform leadfoams with densities as low as 1 g/cm3 using pure leadpowder to which basic lead carbonate was added as ablowing agent. The decomposition range of this blowingagent fits with the melting point of lead. Moreover, by

ll rights reserved.

mailto:[email protected]

4904 A. Irretier, J. Banhart / Acta Materialia 53 (2005) 4903–4917

using a lead compound as the blowing agent one canavoid adding further metals, which could be a usefulproperty in some electrochemical applications wherecontaminations can be deleterious [9]. An example ofsuch a foam can be seen in Fig. 1. This foam was laterused in a battery test.

Lead alloys were also considered. Tin and antimonyare the most important alloying elements of lead. Bothlower the melting temperature of lead and thereforemodify foaming conditions. Moreover, these alloyingelements improve mechanical properties [10] whichmight be important for foam applications.

In addition to technological development it is impor-tant to improve knowledge about the physics of foamingof metals. We therefore investigated the foaming processof lead alloys as a function of time and temperature in aquantitative way and analysed both the kinetics offoaming and the structure of the resulting foams. Alloycomposition and blowing agent content were varied,which allows conclusions to be drawn about the mecha-nisms governing foaming.

The paper is organised as follows: Section 2 describesthe parameter optimisation process which finally led to a

Fig. 1. Application example for lead foam: conductive electrode gridin a lead-acid battery, inset: magnified view of a foam section cut byEDM (pure lead, foamed with 2 wt.% (PbCO3)2 Æ Pb(OH)2 at 450 �C).

formulation yielding lead alloy foams of good quality.After a description of experimental techniques in Sec-tion 3, Section 4 then contains some investigations ofthe foaming behaviour of lead alloys based on the opti-mal procedure. Is it inevitable that the search for goodprocessing parameters implies some preliminary inter-pretation, which has to be given already in Section 2.However, the final discussion and conclusions are pre-sented in Section 5 in the light of all experiments.

2. Establishment of process parameters

A first task was to establish an experimental proce-dure for producing stable lead foams. Preliminary exper-iments failed in which aluminium powder was simplyreplaced by lead and pressing temperatures were re-duced accordingly, while otherwise the same blowingagents and pressing procedures known for aluminiumalloys were applied. The resulting compacted precursormaterial could not be foamed by melting. It becameclear that the selection of both metal powder and blow-ing agent and also the choice of a suitable compactionmethod and the associated adjustment of pressingparameters had to be carried out afresh for lead. Expe-rience with foaming aluminium alloys, however, wasuseful since it helped us in pre-selecting promising pro-cess parameter combinations.

2.1. Variation of materials and process parameters

2.1.1. Metal powder

Twelve different lead powders were purchased cover-ing a variety of product types, nominal powder sizeranges and purities. Size distributions and oxide con-tents were determined by laser particle size analysis(Coulter LPA) and hot gas extraction (LECO TC436),respectively. The parameters characterising these pow-ders as given by the manufacturer and as measured byus are given in Table 1. The measured oxygen contenthas been converted to a volume content of the oxidePbO which is stable at high temperatures (density:9.64 g/cm3 [11]) because this property is actually moremeaningful in the context of foam stabilisation. ThisPbO content ranges from 1% to 65 vol.% for the differ-ent powders.

In a second step, tin (Alfa Aesar, <150 lm, purity99.5%) and/or antimony powder (Chempur, <100 lm,purity 99%) were added to the lead powder to create bin-ary or ternary alloys from these simple eutectic systems.

2.1.2. Blowing agent

The temperature range in which a blowing agent re-leases gas has to match the melting range of the alloyto be foamed [12]. As the precise decomposition behav-iour of inorganic substances is not always known and

Table 1Properties of the various lead powders evaluated

Manufacturer Product type* Particle sizespectrum* (lm)

Purity*(%)

Mean particlesize (lm)

Oxygen content(wt.%)

Equivalent pbocontent (vol.%)

Foaminess(%)

Goodfellow Powder <150 99.5 56 2.86 47 0ALFA Powder <75 99 n.d. 4.11 65 0Chempur Powder <150 99.5 n.d. 1.21 21 20ALFA Powder <150 99.5 9 1.01 17 23Aldrich Shot <500 99.9 n.d. 0.27 5 37Chempur Shot 400–800 99.9 n.d. 0.06 1 37Eckart Powder <63 99.5 9 0.15 3 43Goodfellow Powder 75–180 99.9 148 0.23 4 47Riedel de Haen Granulate <800 ‘‘pure’’ n.d. 0.03 0.5 50ALFA Powder <150 99.5 55 0.09 1.6 70ABCR Powder <150 99.5 59 0.46 8 80Chemco/Cerac Powder 44–105 99.9 82 0.12 (0.16) 2 (3) 93

Properties given by the manufacturer are marked by an asterisk, all others were measured by the authors (n.d. = not determined). Bold writingindicates the powder finally selected as the best. The entries are sorted by increasing ‘‘foaminess’’ (term explained in Section 2.2.4).

Fig. 2. (a) (PbCO3)2 Æ Pb(OH)2 powder used as blowing agent, (b) leadpowder (large particles) mixed with 5% tin powder (small roundparticles) and 0.5 wt.% of (PbCO3)2 Æ Pb(OH)2 (tiny particles on top ofPb). Both images were obtained using a SEM operated at 20 keV.

A. Irretier, J. Banhart / Acta Materialia 53 (2005) 4903–4917 4905

also might be influenced by the compaction process andthe surrounding metallic matrix we chose a variety ofhydrides (Mg-, Ti-, Zr-) and carbonates (Cu-, Mg, Zn-,Pb-) and carried out trial experiments.

2.1.3. Mixing of powders

Metal and blowing agent powders were mixed in atumbling mixer for 30 min. Quite frequently agglomera-tion of the very fine carbonate particles (see Fig. 2(a))was observed. Adding steel balls to the mixing containerhelped in improving the results. The quality of mixingwas checked by visual inspection which was facilitatedby the difference in colour between the powders (leadpowders are dark grey, most carbonates white). The ab-sence of visible clusters of carbonate was interpreted as agood result. A microscopic view of such a good mixture(Fig. 2(b)) shows large lead particles covered with veryfine (PbCO3)2 Æ Pb(OH)2 particles. The tin particles areseen as small spheroid particles in between. Such mix-tures proved to be well compactable and to yield goodfoams in some cases. No further optimisation of powdermixing was carried out for this reason.

2.1.4. Compaction of powder mixturesA simple and often used method for the compaction

of powder mixtures to dense precursors is hot pressing.Therefore, the first experiments for making foamablelead precursors were carried out in a hot pressing tool.Powder mixtures of pure lead (m.p. 327 �C) and blowingagent were compacted at 250 �C and a pressure of110 MPa for 20 min. Tablets with 29 mm diameter and8 mm height were obtained.

After it was realised that hot pressing did not yielduseful results (see Section 2.2.1) we constructed anextrusion tool aiming at creating a higher degree ofshear in the material. This is thought to improve bond-ing between the individual particles during material flowand to avoid the problems encountered with hot pressed

powders. As shown in Fig. 3 the tool consists of a verti-cally arranged cylindrical die which can be closed at thebottom and which has a round bore of 8 mm diameter atthe side. The die was filled with 350 g of lead powdermixture, after which the ram was inserted and the

Fig. 3. Tool used for lead powder extrusion.

Fig. 4. Top view of hot-pressed precursor sample heated to 450 �C. Nofoaming was observed but liquid lead was squeezed out of tablet.


powder was pre-compacted at 20 kN and room temper-ature. The bore hole was closed with a thin aluminiumfoil at this stage. The filled die was then pre-heated to275 �C and held for 2 h after which final pressing wasinitiated. As soon as the force exceeded about 100 kNthe metal started to flow and a compact wire was ex-truded at a rate of about 10 mm/s. We call this proce-dure transverse extrusion. The extrusion ratio was 1:14in the present case.

2.1.5. Foaming

Simple free-foaming experiments were used for thecharacterisation of foamability. A furnace was pre-heated to and held at a given temperature. A coppersheet was used as a substrate for foaming. Pieces of com-pacted precursor material were placed on the pre-heatedsheet. The foaming process was observed visuallythrough a window in the furnace. After maximumexpansion was reached the foam was taken out of thefurnace, cooled with a ventilator and allowed to solidify.Furnace temperatures ranged from 350 to 550 �C.

2.2. Influence of various parameters on foam quality

We shall now review the results of parameter optimi-sation starting with the most important step, namelypowder consolidation.

2.2.1. Powder compaction

Hot-pressed tablets were heated in a furnace set to450 �C. We observed melting of the precursor and the

occurrence of melted droplets on the surface of the pre-cursor but no foaming as can be seen in Fig. 4.

We could not achieve any satisfactory foaming for avariety of different powders, blowing agents and press-ing parameters. We suspect that the densification ofpowders was not sufficient to shear off the oxide layersfrom the individual particles and to create a metallicbonding and a gas-tight structure. Instead the liquidwas squeezed out by the evolving gas and no bubbleswere formed.

Extrusion as described in Section 2.1.4 immediatelyled to very promising results. Most extruded precursorsshowed some degree of foam expansion. No melt drop-lets were formed at the surface in most cases. Expansionwas smooth and the achievable expansion factors ex-ceeded 11 in the best cases. Many of the foams had auniform pore structure such as the one shown inFig. 1. For this reason we did not carry out any furtheroptimisation of powder consolidation. Relative densitiesof the extruded bars were above 98.2% for pure lead andabove 96.6% for all the alloys investigated. There is noclear dependence of the final density on alloycomposition.

2.2.2. Blowing agent

We obtained the best results using lead carbonatewith the nominal composition (PbCO3)2 Æ Pb(OH)2(99%, Alfa Products). This carbonate is called basic leadcarbonate or white lead. Its mean particle size was deter-mined by laser particle analysis. However, the value of11 lm which was obtained does not seem to representthe true particle size if one compares it to the apparentsize of most particles in Fig. 2(a). This is an indication


of strong agglomeration which was not removed even bythe ultrasound-assisted dispersion in the particleanalyser.

Use of (PbCO3)2 Æ Pb(OH)2 led to said volume expan-sion factors up to 11. The blowing agent content wasvaried from 0.5 to 3 wt.%. It was found that increasingblowing agent contents exacerbate the problems withmixing and compaction. 2 wt.% were chosen as a goodcompromise. Surprisingly a similar water-free carbonatePbCO3 did not lead to good foam expansions. Secondbest was MgH2, yielding a volume expansion factor ofslightly above 2.

2.2.3. Foaming temperatureFurnace temperature clearly influences foaming as we

know from aluminium foams [13]. Lower foaming tem-peratures prolong the time needed to reach the meltingrange and also reduce the degree of final foam expansionbecause during slow heating gas can escape from theexpanding precursor without creating pores. Fig. 5shows samples foamed at two different temperatures.A pronounced difference in foam expansion can be no-ticed. Above 500 �C foam expansion tends to reach alimit while oxidation becomes very strong. 450 �C was

Fig. 5. Lead foams produced at 400 �C (upper) and 500 �C (lower).8 mm thick extruded wires containing 2 wt.% (PbCO3)2 Æ Pb(OH)2were used as precursors.

Table 2Quality assessment of lead foams

Foam property Weight factor Points

0 1

Surface quality 1 Very jagged Uneven, many hPore structure 4 No pores Mainly open porExpansion factor 3 <1 >1

<2Strength of foam 2 Very fragile Low

The four properties given in the left column were valued with up to three poinand then summed up a total quality number. A ‘‘foaminess’’ is then defined(explained in Section 2.2.4 and given in Table 1).

therefore taken as a ‘‘standard temperature’’ for com-parative tests. The foaming process is very fast at thistemperature. After the onset of melting, foam expansionup to the maximum volume can take place in the orderof 10 s depending on sample size.

2.2.4. Metal powder

To be able to assess the suitability of a given leadpowder we defined an evaluation system for the result-ing foams [14]. Four properties were evaluated and 0to 3 points given for each property, each weighted bya factor expressing the importance we assign to thisproperty guided by our experience. The points were thensummed up to yield a final quality number (see Table 2),the highest possible value of this number being 30. Thepercentage of these 30 points which was actually reachedfor a given lead powder is called foaminess and is givenin Table 1 for all the materials considered. We foundthat the most important parameters influencing foama-bility were oxygen content and powder size. The best re-sults were obtained using a powder from Cerac-Chemcowith an average particle size of 82 lm (90 vol.% >50 lm,d50 = 80 lm, 90 vol.% <141 lm). This powder reached28 points, e.g., 93% of the possible 30 points. It hadgood flow properties, a fairly light grey colour and didnot agglomerate. The oxygen content of the freshly ob-tained powder packed under argon was 0.12 wt.%. Afterbeing stored in a glass container under air the originalcontent increased up to 0.16 wt.% after one week. Sixmonths after the measurements the powder was charac-terised once more and then had an oxygen content of0.37 wt.% [15].

Fig. 6(b) shows an example for a foam based on thispowder.

Powders which did not lead to good foaming fall intotwo categories. The first includes powders with a highoxygen content up to 4 wt.% which did not flow welland had a lumpy consistency and a dark grey or evenreddish colour. The extruded precursor did not foamwell and showed a similar behaviour already observedfor hot-pressed tablets, i.e., the formation of melt drop-lets at the surface. These powders obtained only 20% ofthe possible points.

2 3

oles Some defects Smooth and closedosity Closed pores, but irregular Closed pores, uniform size

>2 >3<3Medium High

ts each which were then first weighted with the factor given in column 2as the percentage of the possible 30 points which is actually reached

Fig. 6. Foams prepared from three different powders [16]: (a) Chempurlead shot with 0.06 wt.% oxygen, (b) Cerac-Chemco powder with0.16 wt.% oxygen, (c) Chempur powder with 1.21 wt.% oxygen.


Fig. 6(c) shows an example for a ‘‘foam’’ made fromthese materials.

Powders with a very low oxygen content down to0.06 wt.% were spherical and quite coarse. They wereeasy to compact and showed some foaming. However,the foams were not stable. Strong drainage occurredduring foaming and a pool of liquid metal emergedat the bottom of the foam. 30–50% of the possibleassessment points confirmed the inferiority of thesepowders.

Fig. 6(a) shows a foam made with one such powder.As the oxygen content of lead changes gradually dur-

ing exposition to air due to reactions with O2 or CO2 wechecked for possible changes during processing. Leadpowder without lead carbonate additions was movedin the mixer and then extruded. The oxygen contentsof the starting powder, the ‘‘mixed’’ powder and the ex-truded bar were then measured. We found the same va-lue for all samples within a limit of 20%. Therefore, theoxygen content remains nearly constant during process-ing. Long exposure to air at room temperature, how-ever, leads to progressive oxidation or carbonateformation as mentioned above.

2.2.5. Alloy composition

In order to tailor foam properties tin and antimonywere added to the lead powder. Processing parameterswere not changed except that the pressing temperaturewas adapted to the lower melting temperatures of thePb–Sn , Pb–Sb and Pb–Sb–Sn alloys. It was ensured thatthe pressing temperature was well below the solidus tem-perature of the alloy processed and below the lowestmelting point of any of the elementary components aswe are processing powder mixtures. Alloys with up to30 wt.% Sn and up to 10 wt.% Sb and a ternary alloycontaining 10 wt.% Sn and Sb were successfully pro-cessed to foams. Precursors containing Sb were hardand difficult to cold-roll, alloys containing Sn very soft.The foaming behaviour of these alloys will be describedin Section 4 in a quantitative way.

3. Sample preparation and foam characterisation

technique

3.1. Manufacture of foamable precursor for foaming tests

Powder mixtures reflecting the composition of purelead and the alloys Pb–Sn (5, 10, 30 wt.% Sn), Pb–Sb(3 and 10 wt.% Sb) and PbSb10Sn10 were processed to8 mm thick wires by transverse extrusion. Some of thesewires were cold-rolled to 1.2 mm thickness and then cutto square pieces 35 · 16 mm2 in size to be used in free-foaming experiments.

3.2. Free-foaming experiments and foam characterisation

In order to determine the evolution of pore structureduring the foaming of lead alloys, so-called ‘‘free-foam-ing’’ experiments were carried out. The procedure wassimilar to that described in Section 2.1.5. Samples pre-pared by rolling were placed onto a pre-heated substrateand were allowed to foam for a given time, after whichthe foaming process was interrupted by air quenching.The densities q of the foamed samples were determinedby buoyancy measurements. The foams were then sec-tioned and analysed. For this the specimens wereembedded in a white resin and polished with a grid800 paste. This led to a clear contrast between cell wallsand voids. Images were scanned and analysed quantita-tively using the computer program ‘‘Optimas 6’’. Insome cases artefacts stemming from cell walls inter-rupted by polishing or by pieces of resin that had fallenout of small pores had to be corrected manually. Poresize distributions and the mean apparent pore diameterd were derived by the computer program. As quantita-tive analysis of very small pores turned out to be unreli-able we only took account of pores with a diameterabove 0.5 mm. This of course limits the significance ofthe mean value d. Whenever d is close to 0.5 mm the


calculated d is systematically larger than the true aver-age including all pores. We therefore overestimate d

for foams with a fine porosity and therefore cannotuse d as an absolute measure but rather as a qualitativeindicator for pore size.

3.3. Foaming experiments in expandometer

Free-foaming experiments yield quick informationabout the overall expansion behaviour, but a precise pic-ture especially of the early expansion stages requires aquantitative in situ characterisation method of expan-sion as provided by the so-called expandometer. Theset-up of the expandometer is shown in Fig. 7.

The furnace (not shown) heats a foamable sample (2)placed in a cylindrical mould (1) allowing for verticalexpansion only. The mould is mounted on a fixed sup-port (7). Expansion is monitored by a dual position sen-sor (3,4) measuring the difference between the position

Fig. 7. Set-up of mechanical ‘‘expandometer’’ for measuring foamingkinetics. 1: steel tube, 2: foamable precursor, 3: piston, 4: positionsensor, 5: thermocouple for sample, 6: thermocouple for furnace, 7:sample support.

of the upper piston relative to the sample substrateand thus compensating for most of the thermal expan-sion effects in the experimental set-up not stemmingfrom the sample. Both sample and furnace temperaturesare measured with thermocouples (5,6). A discussion ofthe expandometer can be found in [13]. It was found thatthe weight of the piston has little effect on foam expan-sion in a way that final expansion is only marginally lar-ger in free-foaming experiments than in expandometertests. Foam evolution can be monitored in a reliableway for the expansion phase, while the collapse phaseis often disturbed by frictional effects between pistonand tube.

Specimens for foaming were obtained by cuttingsmall chips off a piece of extruded precursor wire andcompacting them at 200 kN to cylindrical tablets of29 mm diameter (at room temperature). Direct hotpressing of powders to tablets was not possible due tothe limitations described in Section 2.1.1. For eachparameter set two specimens were characterised.

3.4. Thermoanalysis

A Netzsch 409C simultaneous thermal analyser wasused to measure heat flow and mass change during con-tinuous heating of blowing agent powder. The heatingrate was 10 K/min and the argon flow rate 50 ml/min.100 mg of powder was characterised in alumina cruci-bles in each experiment.

4. Results

In this section, we describe investigations related tothe foaming process of lead and lead alloys. After ananalysis of the blowing agent and the foamable precur-sor the expansion kinetics of foaming lead alloys areinvestigated either in free-foaming experiments or inthe expandometer.

4.1. Thermo-analysis of blowing agent

Fig. 8 shows thermo-analytical data for (PbCO3)2 ÆPb(OH)2. In the DSC trace we can distinguish two dis-tinct decomposition stages, marked ‘‘I’’ and ‘‘II’’. Thefirst takes place from about 266 (determined by the tan-gent method) to a maximum around 309 �C, the secondfrom 323 to 359 �C. In addition, there is a small endo-thermal reaction below 100 �C. Corresponding to theDSC trace the mass loss also takes place in two stagesas can be seen best by considering the derivative of theTGA curve which matches the DSC trace very well.The first peak of this mass change curve is more pro-nounced with an area 2.25 times that of the second.Therefore, main gas evolution occurs slightly belowthe melting point of pure lead which is indicated in

Fig. 9. Sections of cold rolled (a) PbSn10, (b) PbSb10 precursormaterial containing 2 wt.% (PbCO3)2 Æ Pb(OH)2 (optical microscopy).

Fig. 8. DSC and TGA trace of blowing agent (PbCO3)2 Æ Pb(OH)2.Heating rate 10 K/min, argon atmosphere. Mass change Dm (TGA),negative derivative of Dm and heat flow (DSC) are shown. The scale of�d/dT Dm was chosen such that the height of the first peak matchedthe corresponding DSC peak. Vertical line: melting point of lead.


Fig. 8 as a short vertical line. The lead alloys investi-gated have melting ranges which are below the maxi-mum of the first decomposition peak but still fall intothe decomposition range.

4.2. Microstructure of precursor material

Fig. 9 shows cut sections of two foamable lead alloyprecursors after rolling. Polishing the lead specimensturned out to be difficult. Especially for PbSb10 theintegrity of the rolled material was not sufficient tomaintain a dense matrix. Particles fell out of the surfaceto be prepared leading to some porosity (see Fig. 9(b)).Therefore, the samples could not be fully polished to re-veal microstructural details in depth. What is clearly dis-cernible is that Sb particles embedded in lead maintaintheir shape even after rolling, whereas Sn particles arelargely elongated in the rolled state (see Fig. 9(a)). ThePb–Sn system was less prone to the difficulties encoun-tered with Pb–Sb and showed a better structural integ-rity and consequently fewer pores.

4.3. Free-foaming of lead and lead alloys

4.3.1. Influence of foaming time and alloy composition

Fig. 10 compares the density variation with foamingtime for different lead and lead alloy foams. Here foam-ing time means the time elapsed between the first contactof the foamable precursor with the sample substrate andthe initiation of sample cooling. The time span shown inFig. 10 ranges from 10 to 60 s. Shorter foaming timeswere difficult to control and were therefore not consid-ered. The density drops during the expansion phase ofthe foam, reaches a minimum and then either stays con-stant or rises again indicating some foam collapse. Obvi-ously foaming is a strongly non-linear process. In the

first 10 s not resolved here most of the expansion takesplace. After 60 s the changes are quite small. Pure leadreaches a minimum density after about 20 s and is quitestable after. Additions of antimony lead to both a higherminimum density and a more pronounced collapse aftermaximum expansion. This effect is stronger for 10% Sbthan for 3%. Additions of tin, on the other hand, en-hance foaming and not only produce lower final densi-ties but also reduce collapse. After maximumexpansion density remains almost constant up to thelongest foaming time. The ternary alloy PbSb10Sn10 be-haves very much like binary Pb–Sn alloys.

The evolution of pore structure during foaming canbe monitored by comparing sections of samples foamedfor different times.

Fig. 11 shows a series of such images for thePbSb10Sn10 alloy. Corresponding to volume expansion,the density q decreases with foaming time while theaverage equivalent pore size d increases. The distribu-tion of pore sizes and shapes in the foams is quite irreg-ular, especially for the longer foaming times where manysmall pores co-exist with only a few large pores with

Fig. 10. Density of various lead and lead alloy foams as a function offoaming time. Seven different alloys were investigated. All samplescontained 2 wt.% (PbCO3)2 Æ Pb(OH)2. Foaming temperature was450 �C. t = 0 corresponds to the first contact of the precursor withthe pre-heated substrate. Data points are connected by guidelines tovisualise the general course of foaming.


diameters up to 4 mm. The mean apparent pore diame-ter d ranges from 0.7 to 0.95 during the time intervalconsidered. The true change in pore size is underesti-mated by our analysis as explained in Section 3.2.

4.3.2. Influence of blowing agent content

A series of different alloys containing either 0.5 or2 wt.% blowing agent (PbCO3)2 Æ Pb(OH)2 were foamedat 450 �C. Foaming times ranged from 10 to 60 s. Thedensities of the resulting foams are shown in Fig. 12 asa function of time.

Fig. 11. Sequence of foaming steps in PbSb10Sn10 containing 2 wt.% (PbCOapparent pore diameter d are given.

Foam expansion – reflected by lower foam densities –is much stronger for the higher content of blowingagent. Moreover, expansion is faster. Alloys containing2% blowing agent complete their expansion after 12–20 sand reach minimum densities below 1 g/cm3 for fully ex-panded PbSn5 and PbSn10 alloys. Foams with 0.5%blowing agent continue expanding slowly even after30 s (except for PbSb3 where they collapse). Pure leaddoes not reach densities below 2.4 g/cm3 even after60 s of foaming. PbSb3 expands quickly up to a mini-mum density of 2.3 g/cm3 but then collapses. Only thealloys containing tin reach lower densities around1.7 g/cm3 and appear stable within the first 60 s.

Images of foams based on pure lead and two Pb–Snalloys after 60 s of heating are compared in Fig. 13.For the foams blown with 2% blowing agent, cells aremore expanded. Densities are between 40% and 55%lower than for 0.5% blowing agent. There is also amarked difference in pore size. 2% blowing agent con-tent produces quite a number of large cells and an aver-age cell size d which is nominally 33% and 64% higherthan for 0.5% blowing agent content. This difference iseven larger in reality as explained in Section 3.2 sinced for the foams made with 0.5% blowing agent is quiteclose to the cut-off size of 0.5 mm in our analysis.

4.3.3. Influence of foaming temperature

Fig. 14 gives the density evolution of various lead al-loy foams for a lower foaming temperature than that inFig. 10, namely 400 �C. Obviously, foaming is slower for400 �C than for 450 �C. At a given time, e.g., 12 s, den-sities are higher for most alloys than the densitiesreached at 450 �C. Densities corresponding to maximum

3)2 Æ Pb(OH)2. Foaming temperature was 450 �C. Density q and mean

Fig. 12. Density of various lead and lead alloy foams as a function of foaming time. Comparison of blowing agent contents: 0.5 and 2 wt.%(PbCO3)2 Æ Pb(OH)2. Foaming temperature was 450 �C. Curves for 2% blowing agent as given in Fig. 12 are represented by thin lines. Data points for0.5 wt.% blowing agent are connected by guidelines to visualise the general course of foaming.

Fig. 13. Pore morphology of foams after 60 s of foaming at 450 �C. Two different contents of blowing agent: upper – 0.5 wt.%; lower – 2 wt.%.Densities q and mean apparent pore diameters d are given.


Fig. 14. Same as Fig. 12 for foaming temperature 400 �C.


expansion, however, are only slightly higher for 400 �C.One difference is stability of some of the foams: foamingat the lower temperature slightly reduces the degree ofcollapse of the two Pb–Sb alloys.

Fig. 15 shows the influence of foaming temperatureon (a) pore size, and (b) density. A constant foamingtime of 16 s was chosen, because at this time most foamsshown were still quite stable and little collapse had oc-curred. Pb–Sb foams were not included in Fig. 15(a)since they exhibit very pronounced collapse phenomenaat higher temperatures (even more than the collapseshown in Figs. 10 and 12). Pore size values obtained

Fig. 15. (a) Mean apparent pore diameter of samples, and (b) densityof samples foamed at different temperatures for 16 s (2 wt.%(PbCO3)2 Æ Pb(OH)2).

for these alloys showed an erratic behaviour so thatmerely their density is given in Fig. 15(b). A first obser-vation is that higher foaming temperatures lead to largerpore sizes for all except one data point. Moreover, thereis a clear dependence of pore size evolution on alloy con-tent. Pure lead develops the coarsest pore structure.Increasing tin content leads to smaller pores. The ter-nary alloy PbSb10Sn10 lies between PbSn10 andPbSn30.

The average densities of the foams, see Fig. 15(b),show a pronounced dependence on temperature too.All samples expect PbSb3 and PbSn5 exhibit a densityminimum between 450 and 500 �C. The lowest densityof all samples is found for PbSb10Sn10 foamed at500 �C. This foam not only has a fine pore structurebut also a density of only 0.93 g/cm3, which correspondsto an expansion factor of 12.2.

4.4. Expandometer foaming

4.4.1. Influence of sample thickness

In a first step we determined the influence of sampledimensions on the foaming behaviour of pure lead pre-cursors. In Fig. 16 the expansion kinetics of pure leadfoams and the corresponding sample temperatures aredisplayed. Expansion is given as g = V/V0, where V isthe current foam volume, V0 the volume of the unfoa-med precursor. All samples were cylindrical, with aconstant diameter of 29 mm, but sample height wasvaried. The temperature curves show that the heatingrate in the thinnest sample is highest, followed by thethicker specimens. In all cases the arrest point duringheating is close to 327 �C which is the melting pointof lead. After melting the temperature in the samplesfurther rises and approaches the furnace temperature.The corresponding expansion curves show that thethinnest sample starts to expand first, while expansionof the thicker specimens is delayed, reflecting delayed

Fig. 16. Expansion kinetics of pure lead precursor samples with threedifferent thicknesses. Blowing agent: 2 wt.% (PbCO3)2 Æ Pb(OH)2,furnace temperature 450 �C.


heating. Foaming starts as soon as the melting point oflead has been reached. The overall foaming course isthe same in all three cases. A maximum of foamingis observed at about 350 �C after which a regime ofcollapse follows. Collapse is more pronounced in theexpandometer than in free-foaming under equal condi-tions (compare to Fig. 10). The thicker sample reachesa lower maximum expansion which can easily be attrib-uted to the higher sample mass resting on the expand-ing foam and hindering expansion. The two thinnersamples, however, are not too different concerning theirmaximum expansion. Therefore, it can be concludedthat precursor thicknesses up to 6 mm lead to the sameresults and 6 mm was selected as standard thickness forall the expandometer experiments. This corresponds toa sample weight of about 40 g.

4.4.2. Influence of alloy composition

The foaming behaviour of some Pb–Sn and one ter-nary alloy was compared to pure lead in a series ofexpandometer experiments. Fig. 17 shows the corre-sponding expansion diagrams. Obviously, addition oftin shifts the onset of foaming to earlier times by100 s in Fig. 17(a). In addition, maximum foam expan-sion increases with tin content, a notable increase beingobserved for 10% Sn or more. All foams show somecollapse after maximum expansion. The ternary alloyPbSb10Sn10 expands slightly more than PnSn10 butfoam collapse is a bit more pronounced. In contrastto alloys containing Sn, alloys with Sb (Fig. 17(b)) ex-

Fig. 17. Expansion kinetics of pure lead and some alloys containing(a) Sn and (b) Sb. Blowing agent: 2 wt.% (PbCO3)2 Æ Pb(OH)2, furnacetemperature 450 �C.

hibit slightly lower maximum expansions than purelead while foam collapse appears more pronounced.The onset of foaming is also shifted to earlier timesby Sb additions.

Analysing the temperature at which foam expansionbegins – defined as the point at which expansion reaches1% of maximum expansion – we find that pure leadstarts at 327 �C which is exactly its melting point. Addi-tions of Sb lower the onset of expansion to about 305 �Calmost independently of concentration. Additions of Sn,on the other hand, lead to an onset temperature around250 �C without much dependence on tin concentration.The same applies to the ternary alloy PbSb10Sn10 whichstarts at 251 �C.

5. Discussion

5.1. Metal powder properties

The experience that successful foaming of lead alloysrequires powder compaction by transverse extrusion andthe found sensitivity of foaming behaviour on powderproperties shows that the surfaces of the powder parti-cles, namely, the oxide or carbonate layers on individualparticles play a very important role. During compactionof metal powder mixtures the oxide layer of each indi-vidual powder particle has to be sheared off to allowfor the creation of a metallic bonding between neigh-bouring powder particles and to ensure that blowingagent particles are enclosed in a dense metallic matrix.Hot compaction, sufficient for making foamable precur-sors from aluminium alloys, does not provide enoughshear to densify lead powders, at least not the commer-cial powders used. As a consequence, gas escapes fromthe precursor without creating bubbles and merelypushes out some of the liquid during heating. This givesrise to the often observed droplets as shown in Fig. 4.For too high oxide contents even transverse extrusionis not sufficient to break up the oxide layers and no foa-mability can be achieved at all (Fig. 6(c)). Reducing theoxygen content to a very low level is not a good strategyto improve this situation since, on the other hand, a cer-tain oxide content is necessary to ensure foam stabilityas shown in Fig. 6. It is well known that metal foamsare stabilised by solid particles floating in the melt[17–19], a phenomenon which is also known from aque-ous foams and emulsions [20]. Wubben could recentlyshow [15] that the ratio between the total internal poresurface of a lead foam is proportional to the volumecontent of the entrained oxide and concluded that theoxides float on the foam films during foaming and createstabilisation by a still disputed mechanism [21]. Conse-quently, a too low oxide content leads to coarsening un-til the pore surface area has adjusted to the availableamount of oxide.


The best choice of lead powder is a compromise be-tween various requirements. This is reflected by theproperties of our best powder, a medium coarse powderof 82 lm mean particle size and a medium oxygen con-tent, i.e., in the order of 0.2 wt.%, corresponding to3 vol.% PbO. From the difficulties encountered it be-comes clear that metal powder properties play a moreimportant role for lead than for aluminium in a sensethat the quality of the emerging foam depends sensi-tively on the nature of the powder used.

Aluminium foams contain about 1 vol.% Al2O3 whichis sufficient to stabilise the films [13,22]. Therefore, theoxide content in lead foams is higher than in aluminiumbut has the same order of magnitude. We therefore as-sume that the basic stabilisation mechanism is the same.In addition to the oxides originating from the formerpow-der surfaces one has to take account of the residues fromthe blowing agent which are present in lead foams (leadoxide) and aluminium foams (titanium or titaniumoxide). These residues could also contribute to stabilisa-tion, although their effectiveness is probably low becausethese particles are quite coarse compared with the submi-crometer size of the oxides present in the precursor.

Addition of alloying elements has an important im-pact on both structure and properties of the precursor.The differences observed between Sn and Sb are mainlydue to the difference in hardness of these elements. TheBrinell hardness 3.9 of tin is just slightly higher than thatof Pb (2.4), whereas Sb has a much higher hardnessabove 30 [23]. This explains why only tin particles aredeformed during extrusion, whereas Sb particles remainunchanged. The intense and simultaneous deformationof both tin and lead can be thought to yield the betterstructural integrity observed for extruded wires madefrom Pb–Sn alloys.

5.2. Blowing agent

In order to ensure good foaming the decompositioncharacteristics of the blowing agent have to match themelting range of the alloys to be foamed [12]. Titaniumand zirconium hydrides successfully used for foaming Alor Zn alloys did not yield good results because theirmain decomposition takes place above about 400 �C[24,25]. The decomposition behaviour of the most suit-able blowing agent found in this study as shown inFig. 9 clearly meets the requirements more than thesehydrides. However, the reason why MgH2 and othercarbonates studied led to inferior results, although theirdecomposition takes place in the required temperaturerange, is not entirely clear. It has to be assumed that be-side the temperature dependence of gas release fromloose powder other factors play a role such as the waythe blowing agent particles are embedded in the metallicmatrix during extrusion. This might depend on morpho-logical properties of the powder particles and give rise to

a different residual porosity or structural integrity of theprecursor which in turn influences foaming.

Given the decomposition reaction of the blowingagent and the known molar weights of the componentsinvolved:

we can calculate the weight reduction of the precursorafter total decomposition. This theoretical reduction is�13.7% which roughly corresponds to the measuredmass decrease of �16.5% (see Fig. 8).

We can calculate the maximum gas volume which canbe released from a given volume of foamable precursor.Accordingly, 1 cm3 of precursor containing x wt.% of(PbCO3)2 Æ Pb(OH)2 releases 7.5/x cm3/wt.% of waterand 15/x cm3/wt.% of CO2 at the m.p. of lead (327 �C).In order to achieve a maximum expansion factor of,e.g., 11 which is actually observed when foaming precur-sors containing 2 wt.% blowing agent (corresponding toa final foam density of 1 g/cm3) one would need 10 cm3

of gas from each cm3 of precursor. If water and CO2

acted in the same way and no gas was lost, a blowingagent content of x = 10/(7.5 + 15) � 0.45 wt.% shouldyield this expansion factor. In reality we observe thatno full expansion is reached at this level of blowing agentas can seen from Fig. 12. Therefore, either gas is lost dur-ing foaming and/or decomposition is incomplete at thetemperatures applied. Fig. 9 actually shows that onlyabout 2/3 of the total gas amount is released at them.p. of lead and even less for lead alloys. Therefore,gas losses can be inferred, which is plausible since the pre-cursor is not totally dense and the surface of the emergingfoam is rough and contains holes (Fig. 5). A blowingagent content of 2 wt.% seems to be the ideal choice bothfrom the theoretical and the experimental viewpoint.

Increasing the blowing agent content beyond the levelof 2 wt.% created problems with mixing and compactionas pointed out already. Even if it were possible to usemore blowing agent it is not sure whether this would leadto lower densities. Comparing foaming for two differentblowing agent contents in Fig. 12 we see that maximumexpansion is not four times higher for 2% compared to0.5%. In aluminium foams expansion does not further in-crease beyond a certain level (about a factor 5 for a com-mon alloy such as AlSi7) if the blowing agent content isincreased [26]. Instead, cells coalesce as films are over-stretched and the foam is prone to collapse [27]. Fig. 13shows that pore morphologies depend very much onthe blowing agent content. The higher content of2 wt.% leads to a lower final density after 60 s of foamingbut at the cost of a more heterogeneous pore structure.Obviously the stronger expansion has triggered ruptureevents, a phenomenon called growth coalescence by


Korner et al., who showed that density and pore size arecorrelated [28]. Our data does not show this clear corre-lation – perhaps due to the low precision of pore sizemeasurement – but it shows the existence of this effect.

5.3. Foaming process

The time dependence of foam expansion as ex-pressed by the expandometer tests looks very muchlike that of aluminium alloys (compare expansion dia-grams, e.g., in Fig. 17, with those in [13]). Foamexpansion starts as soon as an internal gas pressureis building up and enough liquid is available to allowfor the evolution of liquid films. As the furnace tem-perature lies above the liquidus temperature of the al-loy to be foamed, there is always some overheating inlate stages of foam expansion and correspondinglycollapse phenomena are observed. The evolution ofpore structure as seen in Fig. 11 also resembles thatof aluminium. We observe a growth in size and areduction in the number of cells, i.e., coalescence dri-ven by expansion.

The various foaming experiments show that thefoaming behaviour of lead alloys depends on the sameparameters known from foaming aluminium, namelyfoaming temperature and alloy composition.

Foaming temperature influences both foaming kinet-ics and expansion behaviour. Obviously, higher foam-ing temperatures lead to faster heating up to themelting temperature, more efficient heat transfer intothe melting material and therefore earlier foaming. Acomparison between Figs. 10 and 14 shows this veryclearly. At higher temperatures, i.e., at T P 500 �C,higher foam expansions can be achieved compared tolow temperatures if foaming is stopped at maximumexpansion as demonstrated in Fig. 5. For lower temper-ature, however, the difference is quite small (compare400 and 450 �C in Figs. 12 and 14). For a fixed foamingtime, e.g., 16 s, the picture is more complex (seeFig. 15(b)). Here the lowest densities are achieved atintermediary temperatures for most alloys. This meansthat an increase in temperature first leads to a highervolume expansion (except for PbSn5 and especiallyPbSb3), later to a reduction in expansion (all alloys).Therefore, an optimum temperature for highest foamexpansion exists at a given time. Higher temperaturesthan this optimum lead to a more pronounced foamcollapse and higher densities, lower temperatures toretarded or incomplete foaming. The temperaturedependence of collapse is best seen for Pb–Sb alloysin Figs. 10 and 14. These figures show a certain degreeof collapse after maximum expansion in all cases, butcollapse is much stronger for 450 �C than for 400 �C.The reason for this must be the lower viscosity of meltsat higher temperatures which leads to instable foams iffoam stabilisation is insufficient. Higher temperatures

also lead to coarser pore structures for all alloys asdemonstrated in Fig. 15(a). At higher temperatures(i.e., T P 500 �C) pores become larger although densityincreases. This is another indication for an increasingrate of coalescence in the foam at higher temperatures.

The various alloys investigated behave in quite a dif-ferent way. Taking pure lead as a reference system addi-tions of tin and antimony were investigated. Both lowerthe melting temperature. Consequently, foaming startsearlier as can be seen best from the expandometer mea-surements in Fig. 17. However, the action of Sn and Sbis quite different. Sn improves foaming, i.e., leads tohigher expansions, more stable foams and also to finerpore structures (see Fig. 15(a)), whereas Sb leads topoorer foaming, i.e., foam expansions are lower andfully expanded foams show a pronounced collapse. Notethat the expandometer tests in general show more foamcollapse than free-foaming experiments which can beattributed to the weight of the piston.

As both Sn and Sb lower the melting point of lead,the change in foamability upon alloying cannot be aneffect of melting temperature alone. The processes dur-ing melting of Pb–Sn and Pb–Sb must be responsiblefor the difference. As the ternary alloy PbSb10Sn10shows the favourable behaviour of binary Pb–Sn al-loys a beneficial action of Sn can be inferred. Wethink that the low melting point of tin at 232 �C incontrast to antimony which melts at 631 �C is respon-sible for the difference. While heating up foamablePb–Sn or Pb–Sb–Sn precursors (which are heteroge-neous mixtures of elementary powders) some metallicmelt is formed in a very early stage. This will certainlyspeed up the alloying process and also close the net-work of pressing porosity in the precursor throughwhich part of the blowing gas can escape. In Pb–Sbthe Sb powder particles remain solid until diffusionacross the interfaces between Pb and Sb creates someeutectic liquid. As heating rates are very fast in ourexperiments this is not expected to happen below themelting point of lead at 327 �C. Therefore, only alloyscontaining elementary tin show this behaviour whichis known to assist foaming, namely the early forma-tion of liquid, while Pb–Sb alloys do not.

Another possible reason for the different foamingbehaviour of Pb–Sn and Pb–Sb might be the differentconsistency of the precursor as seen from the micro-graphs in Fig. 9. The addition of soft tin particles yieldsa precursor which has a high structural integrity,whereas additions of antimony do not. As a conse-quence, during foaming of Pb–Sb more internal de-bonding occurs which facilitates the escape of blowinggas. Finally, one could also speculate that the melt prop-erties of Pb–Sn and Pb–Sb are different, e.g., the valuefor viscosity or the proneness to gas–metal reactionswith CO2 or O2 in the different alloys which could alsoinfluence the foaming process. However, the explanation


given above seems more likely since the ternary systembehaves like binary Pb–Sn alloys.

6. Conclusions

� Lead and lead alloy foams could be produced bymodifying the powder compact process known foraluminium alloys. Uniform foams with densitiesbelow 1 g/cm3 could be obtained.

� Foaming is very sensitive to the properties of the leadpowder used. Dry and fluid powder with a grain sizeof �80 lm and a moderate oxygen content(�0.2 wt.%) was the best choice.

� Basic lead(II) carbonate (PbCO3)2 Æ Pb(OH)2 wasshown to be a suitable blowing agent.

� Powder mixtures have to be densified by extrusionand can then be foamed ideally at 450 �C.

� The foaming of lead and lead alloys is governed bythe same parameters known to influence foaming ofaluminium alloys. Foaming temperature and alloycomposition are the most important ones.

� Pb–Sn alloys are found to reach higher foam expan-sions and show an increased foam stability as com-pared to pure lead. Pb–Sb alloys are more prone tofoam collapse and drainage. Ternary alloys contain-ing both Sb and Sn behave similar to Pb–Sn foams,i.e., they expand more and are more stable than purelead foams.

� Reasons for the beneficial action of tin are the posi-tive influence on powder compaction and the earlyformation of a liquid phase during heating to thefoaming temperature.

Acknowledgements

The work was supported financially by DLR (Con-tract 50 WM 9821 and 50 WM 0126) and ESA(14308/00/NL/SH). Peter Weferling carried out somepreliminary tests on different lead powders. Accumulat-orenwerke Hoppecke Brilon supported some of the earlywork on lead foams.

References

[1] Banhart J. Prog Mater Sci 2001;46:559.[2] Baumeister J. Verfahren zur Herstellung poroser Metallkorper.

German Patent 4018360; 1990.[3] Allen BC, Mote MW, Sabroff AM. Method of making foamed

metal. US Patent 3,087,807; 1963.[4] Baumgartner F, Duarte I, Banhart J. Adv Eng Mater 2000;2:168.[5] Ashby MF, Evans A, Fleck NA, Gibson LJ, Hutchinson JW,

Wadley HNG. Metal foams – a design guide. Oxford: Butter-worth-Heinemann; 2000.

[6] Degischer HP, Kriszt B, editors. Handbook of cellular met-als. Weinheim: Wiley–VCH; 2002.

[7] Koch H, Burkhardt T. Untersuchungen der supraleitendenEigenschaften geschaumter Bleiproben. Technical Report, Uni-versity of Jena; 2000 [in German].

[8] Wubben T, Stanzick H, Banhart J, Odenbach S. J Phys:Condensed Matter 2003;15:S427.

[9] Banhart J. Verfahren zur Herstellung poroser Metallkorper undVerwendung derselben. German Patent 10115230; 2001.

[10] Hilger JP. J Power Sources 1995;53:45.[11] Lide DR, editor. Handbook of chemistry and physics. Boca

Raton (FL): CRC Press; 2004.[12] von Zeppelin F, Hirscher M, Stanzick H, Banhart J. Compos Sci

Technol 2003;63:2293.[13] Duarte I, Banhart J. Acta Mater 2000;48:2349.[14] Weferling P. Optimierung der Herstellungsparameter von Blei-

schaum. Diploma thesis, Hochschule Bremen; 1999 [in German].[15] Wubben T. Zur Stabilitat flussiger Metallschaume. Forts-

chrittsberichte Series 3, vol. 803. Dusseldorf: VDI-Verlag; 2004[in German].

[16] Banhart J. J Metals 2000;52:22.[17] Ip SW, Wang Y, Toguri JM. Can Metall Quart 1999;38:81.[18] Gergely V, Clyne TW. Adv Eng Mater 2000;2:175.[19] Babcsan N, Leitmeier D, Degischer HP, Banhart J. Adv Eng

Mater 2004;6:421.[20] Binks BP. Curr Opin Colloid Interf Sci 2002;7:21.[21] Kaptay G. Colloids Surfaces A 2004;230:67.[22] Weigand P. Thesis, University of Aachen; 1999 [in German].[23] Available from: www.matweb.com.[24] Matijasevic-Lux B, Rack A, Haibel A, Banhart J. Influence of

powder pre-treatments on metal foam pore structure. In: SingerRF, Korner C, Altstadt V, Munstedt H, editors. Cellular metalsand polymers. Uetikon-Zurich (CH): Transtech Publications;2005. p. 57.

[25] Lehmhus D, Rausch G. Adv Eng Mater 2004;6:313.[26] Stanzick H, Duarte I, Banhart J. Materialwissenschaft Werkst-

offtechnik 2000;31:409.[27] Stanzick H, Wichmann M, Weise J, Banhart J, Helfen L,

Baumbach T. Adv Eng Mater 2002;4:814.[28] Korner C, Arnold M, Singer RF. Mater Sci Eng A 2005;396:28.

http://www.matweb.com

Lead and lead alloy foams - MITweb.mit.edu/3.042/team3_11s/papers/irretier2005.pdfLead and lead...

Documents

Transcript of Lead and lead alloy foams - MITweb.mit.edu/3.042/team3_11s/papers/irretier2005.pdfLead and lead...